structural and functional organization of the synapse

Diversity in Synapse Structure and Composition A synapse has a presynaptic component, usually an axon but sometimes a drite, and a postsynaptic component, usually part of a dendrite, cel

STRUCTURAL AND FUNCTIONAL ORGANIZATION

OF THE SYNAPSE

STRUCTURAL AND FUNCTIONAL ORGANIZATION

OF THE SYNAPSE

Johannes W Hell

University of Iowa Iowa City, IA, USA

Johannes W Hell Michael D Ehlers

Iowa City, IA 52242

DOI: 10.1007/978-0-387-77232-5

Library of Congress Control Number: 2007941249

© 2008 Springer Science+Business Media, LLC

All rights reserved This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY

10013, USA), except for brief excerpts in connection with reviews or scholarly analysis Use in tion with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden

connec-The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject

to proprietary rights

While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made The publisher makes no warranty, express or implied, with respect

to the material contained herein

Printed on acid-free paper

9 8 7 6 5 4 3 2 1

Cover illustration: The cover illustration shows an immunofluorsecence micrograph of a hippocampal

pyramidal neuron at two weeks in culture The neuron was stained for the abundant calcium and calmodulin-dependent protein kinase CaMKII (green) and the presynaptic protein synapsin (red) The pictures was provided by Y Chen and J W Hell, University of Iowa

Duke University Medical Center

Durham, NC 27710

and colleague Alaa El-Husseini

The synapse is a fascinating structure for many reasons Biologically, it is an exquisitely organized subcellular compartment that has a remarkable capacity for fidelity and endurance Computationally, synapses play a central role in signal transmission and processing that represent evolution’s solution to learning and memory Nervous systems, including our own brains, possess an extraordinary capacity for adaptation and memory because the synapse, not the neuron, constitutes the basic unit for information storage Because the molecular complexities underlying signal processing and information storage must occur within the tiny space of the synapse, the precise molecular organization of proteins, lipids, and membranes at the synapse is paramount Given the central role of the synapse in neuronal communication, it comes as no surprise that dysregulation of the synapse accounts for many, if not most, neurological and psychiatric disorders Clinically, the synapse thus constitutes a prime target for treatments of these diseases

It is for these reasons that we have chosen to focus our work on deciphering the structural and functional organization of the synapse We have assembled leaders in the field of synapse biology to describe and distill the wonders and mysteries of the synapse This book provides a fundamental description of the synapse developed over many decades by numerous investigators, paired with recent insight into new aspects of synapse structure and function that is still in flux and at the cutting edge of research This book grew out of a symposium and a research seminar at the University of Iowa that were sponsored, in large part, by the generous support of the Obermann Center for Advanced Studies Obermann Seminars are specifically designed to gather international scholars and produce interdisciplinary research publications

We are grateful for the exceptional efforts of our contributing authors, without whom this book would not have been possible Their willingness to take time from their busy research schedules to share their insight and ideas with the breadth and depth that allow us to compile a collective work that is illuminating and useful, for both the general biologist and specialized neuroscientist, is very much appreciated We express our gratitude to our assistants Ms Susan Harward and Ms Sue Birely for their professionalism and help with the book layout and proof reading Lastly, we would like

to thank our families (Laura and Henrik, Mary, Solon, Anselm, and Hans), who provided the support, encouragement, inspiration and comic relief, that in many ways helped to make this book possible

Michael D Ehlers and Johannes W Hell

Editors

List of Contributors

Diversity in Synapse Structure and Composition 1

Kristen M Harris

The Role of Glutamate Transporters in Synaptic Transmission 23

Dwight E Bergles and Robert H Edwards

Structure and Function of Vertebrate and Invertebrate Active Zones 63

Craig C Garner and Kang Shen

Neurotransmitter Release Machinery:

Components of the Neuronal SNARE Complex and Their Function 91

Deniz Atasoy and Ege T Kavalali

The Molecular Machinery for Synaptic Vesicle Endocytosis 111

Peter S McPherson, Brigitte Ritter, and George J Augustine

Initiation and Regulation of Synaptic Transmission by Presynaptic

Calcium Channel Signaling Complexes 147

Zu-Hang Sheng, Amy Lee, and William A Catterall

Adhesion Molecules at the Synapse 173

Alaa El-Husseini

Dendritic Organelles for Postsynaptic Trafficking 205

Cyril Hanus and Michael D Ehlers

Structure and Mechanism of Action of AMPA and Kainate Receptors 251

Mark L Mayer

Cellular Biology of AMPA Receptor Trafficking and Synaptic Plasticity 271

Jose A Esteban

xiii

Structure and Function of the NMDA Receptor 289

Hongjie Yuan, Matthew T Geballe, Kasper B Hansen, and Stephen F Traynelis

Molecular Properties and Cell Biology of the NMDA Receptor 317

Robert J Wenthold, Rana A Al-Hallaq, Catherine Croft Swanwick,

and Ronald S Petralia

Surface Trafficking of Membrane Proteins at Excitatory

and Inhibitory Synapses 369

Daniel Choquet and Antoine Triller

Scaffold Proteins in the Postsynaptic Density 407

Mary B Kennedy, Edoardo Marcora, and Holly J Carlisle

Ca 2+ Signaling in Dendritic Spines 441

Bernardo L Sabatini and Karel Svoboda

Postsynaptic Targeting of Protein Kinases and Phosphatases 459

Stefan Strack and Johannes W Hell

Long-Term Potentiation 501

John E Lisman and Johannes W Hell

Homeostatic Synaptic Plasticity 535

Gina G Turrigiano

Ubiquitin and Protein Degradation in Synapse Function 553

Thomas D Helton and Michael D Ehlers

Signaling from Synapse to Nucleus 601

Carrie L Heusner and Kelsey C Martin

Molecular Organization of the Postsynaptic Membrane

at Inhibitory Synapses 621

I Lorena Arancibia-Carcamo, Antoine Triller, and Josef T Kittler

Acid-Sensing Ion Channels (ASICs) and pH in Synapse Physiology 661

John A Wemmie, Xiang-ming Zha, and Michael J Welsh

Glia as Active Participants in the Development and Function of Synapses 683

Cagla Eroglu, Ben A Barres and Beth Stevens

Plasticity of Dentate Granule Cell Mossy Fiber Synapses:

A Putative Mechanism of Limbic Epileptogenesis 715

James O McNamara, Yang Z Huang, and Enhui Pan

Stroke – A Synaptic Perspective 731

Robert Meller and Roger P Simon

Neuroplasticity and Pathological Pain 759

Michael W Salter

Index 781

List of Contributors

Rana A Al-Hallaq

Laboratory of Neurochemistry, National Institute on Deafness and Other

Communication Disorders, National Institutes of Health, Bethesda, MD, USA,

e-mail: wenthold@nidcd.nih.gov

I Lorena Arancibia-Carcamo

Department of Pharmacology, University College London, Gower Street, London,

WC1E 6BT, UK, e-mail: l.carcamo@ucl.ac.uk

Deniz Atasoy

Department of Neuroscience, U.T Southwestern Medical Center, 5323 Harry Hines

Boulevard, Dallas, TX 75390-9111, USA, e-mail: ege.kavalali@utsouthwestern.edu

George J Augustine

Department of Neurobiology, Duke University Medical Center, Box 3209, Durham,

NC 27710, USA, e-mail: georgea@neuro.duke.edu

Ben A Barres

Department of Neurobiology, Stanford University School of Medicine, Stanford, CA

94305, USA, e-mail: barres@stanford.edu

Dwight E Bergles

The Solomon H Snyder Department of Neuroscience, Johns Hopkins School of

Medicine, 725 N Wolfe St., WBSB 1003, Baltimore, MD 21205, USA,

Department of Pharmacology, University of Washington, Seattle, WA 98195-7280,

USA, e-mail: wcatt@u.washington.edu

Daniel Choquet

UMR 5091 CNRS, Université de Bordeaux 2, Physiologie Cellulaire de la Synapse,

Institut François Magendie rue Camille Saint Sặns 33077 Bordeaux Cedex, France,

Department of Neurobiology, Stanford University School of Medicine,

Stanford, CA 94305, USA, e-mail: ceroglu@stanfordmedalumni.org

Department of Chemistry, Emory University, Atlanta, GA 30322, USA, e-mail: strayne@emory.edu

Kasper B Hansen

Department of Pharmacology, Emory University School of Medicine,

Atlanta, GA 30322, USA, e-mail: strayne@emory.edu

Thomas D Helton

Department of Neurobiology, Duke University Medical Center, Howard Hughes Medical Institute, Durham, NC 27710, USA, e-mail: helton@neuro.duke.edu Carrie L Heusner

Department of Biological Chemistry and Department of Psychiatry and

Biobehavioral Sciences, Brain Research Institute, Semel Institute for Neuroscience and Human Behavior, UCLA, Los Angeles, CA 90095, USA,

Division of Biology, California Institute of Technology, Pasadena, CA 91125, USA, e-mail: kennedym@its.caltech.edu

Josef T Kittler

Department of Physiology, University College London, Gower Street, London, WC1E 6BT, UK, e-mail: j.kittler@ucl.ac.uk

Amy Lee

Department of Pharmacology, Emory University School of Medicine,

Atlanta, GA 30322, USA, e-mail: alee@pharm.emory.edu

Department of Biological Chemistry and Department of Psychiatry and

Biobehavioral Sciences, Brain Research Institute, Semel Institute for Neuroscience and Human Behavior, UCLA, Los Angeles, CA 90095, USA,

e-mail: kcmartin@mednet.ucla.edu

Mark L Mayer

Porter Neuroscience Research Center, Laboratory of Cellular & Molecular

Neurophysiology, NICHD, NIH, Bethesda, MD 20892, USA,

Laboratory of Neurochemistry, National Institute on Deafness and Other

Communication Disorders, National Institutes of Health, Bethesda, MD, USA,

Brigitte Ritter

Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, 3801 Rue University, Montreal, Quebec, Canada, H3A 2B4, e-mail: peter.mcpherson@mcgill.ca

Department of Neurobiology, Stanford University School of Medicine, Stanford, CA

94305, USA, e-mail: beths@stansfordmedalumni.org

Catherine Croft Swanwick

Laboratory of Neurochemistry, National Institute on Deafness and Other

Communication Disorders, National Institutes of Health, Bethesda, MD, USA, e-mail: wenthold@nidcd.nih.gov

Stephen F Traynelis

Department of Pharmacology, Emory University School of Medicine, Atlanta, GA

30322, USA, e-mail: strayne@emory.edu

Antoine Triller

Inserm UR497, Ecole Normale Supérieure, Biologie Cellulaire de la Synapse N&P,

46, rue d’Ulm 75005 Paris, France, e-mail: triller@biologie.ens.fr

John A Wemmie

Department of Psychiatry, Neuroscience Program, and Department of Veterans Affairs Medical Center, University of Iowa, Roy J and Lucille A Carver College of Medicine, Iowa City, IA,USA, e-mail: john-wemmie@uiowa.edu

Robert J Wenthold

Laboratory of Neurochemistry, National Institute on Deafness and Other

Communication Disorders, National Institutes of Health, Bethesda, MD, USA, e-mail: wenthold@nidcd.nih.gov

Hongjie Yuan

Department of Pharmacology, Emory University School of Medicine,

Atlanta, GA 30322, USA, e-mail: strayne@emory.edu

Xiang-ming Zha

Department of Internal Medicine and Howard Hughes Medical Institute, University

of Iowa, Roy J and Lucille A Carver College of Medicine, Iowa City, IA 52242, USA, e-mail: xiangming-zha@uiowa.edu

Diversity in Synapse Structure and Composition

A synapse has a presynaptic component, usually an axon but sometimes a drite, and a postsynaptic component, usually part of a dendrite, cell soma, or axonal initial segment and occasionally an astroglial process Perisynaptic astroglia repre-sent a third component that occurs at some synapses When perisynaptic astroglia are present, the structural complex has been referred to as a ‘tripartite synapse’ (45) Figure 1 illustrates an electron micrograph of an ultrathin (50 nm) section through dendrites, axons, astroglial processes an synapses in the rat hippocampus This picture was chosen to open this chapter because it nicely illustrates the diversity in the composi-tion of even a tiny segment of the neuropil

den-Fig 1 A single thin section spanning approximately 10 µm2 in the middle of the apical

dendritic arbors of hippocampal area CA1 pyramidal cells (a) Gray scale image obtained at the electron microscope (b) Same section colorized to illustrate dendrites (yellow), axons and

vesicle-filled axonal boutons (green), asymmetric postsynaptic densities of synapses located

on diversely shaped dendritic spines (red), and astroglial processes (lavender) Stubby (s), thin

(t) and mushroom (m) spines can be seen in longitudinal section emerging from dendrites A large mushroom spine head has a perforated (p) postsynaptic density

2 Presynaptic Structures and Composition

Three dimensional reconstructions of individual presynaptic axons that pass through the complex hippocampal neuropil show the diversity in their local trajectories (Fig 2a) A three-dimensional reconstruction of a hippocampal dendritic segment that is approximately 10 microns long illustrates the more than 10 fold variation in the dimensions of neighboring dendritic spines and synapses (Fig 2b)

Fig 2 Diversity in the trajectory of presynaptic axons and shapes of postsynaptic dendritic spines (a) Three-dimensional reconstructions of a subset of axons passing through a single electron micrograph (Adapted from (36)) (b) Three-dimensional reconstruction of a single

dendritic segment, illustrating the diversity in dendritic spine shapes and their postsynaptic

densities (red) Inhibitory shaft synapses are colorized in blue (This is a recently surfaced

image of dendrite 21 from (18) available at http://synapses.clm.utexas.edu/) Scale is mately 0.5 µm3 for both reconstructions

approxi-2.1 Presynaptic Active Zone

A presynaptic axonal bouton contains vesicles with a variety of shapes and sizes These vesicles contain excitatory or inhibitory neurotransmitters, neuromodulatory peptides, proteins required to concentrate neurotransmitters, and a variety of proteins involved in the vesicle cycle or that are destined for the presynaptic active zone (11) The presynaptic active zone is a specialized region variously described as ‘dense projections’, the ‘presynaptic grid’ or mini-active zones (34) where vesicles dock and become ready for release Presynaptic vesicles are arranged along filaments that appear to be connected to the presynaptic active zone area (20, 25)

2.2 Presynaptic Vesicles – Excitatory Synapses

Excitatory presynaptic boutons contain clear round vesicles, approximately 35–50

nm in diameter (Fig 3) These vesicles usually contain the neurotransmitter mate Vesicles docked at the active zones are thought to be ready for release and vesicles located away from the membrane are thought to be in a pool that can be recruited for later release One mechanism for neurotransmitter release involves pore formation and subsequent collapse of the presynaptic vesicle into the presynaptic membrane at the active zone with the contents being released into the synaptic cleft Following exocytosis, synaptic vesicle membrane and protein constituents are recy-cled through endocytosis (see chapter by McPherson et al., this volume) Endocyto-sis typically occurs distant from the active zone, and is characterized morphologi-cally by the presence of clathrin coated pits, vesicles and tubular compartments with coated buds that give rise to new synaptic vesicles locally in the presynaptic bouton (e.g Fig 3g)

gluta-Sorting endosomal complexes also have a multivesicular body (similar to that shown in Fig 12d for dendrites below) or a primary lysosome that transports pro-teins and membrane bound for degradation away from the axonal bouton back to the soma Presynaptic vesicles can also be rapidly recycled through a ‘kiss-and-run’ mechanism During kiss-and-run, the vesicles release a portion of their contents through the pore, without collapse of the vesicular membrane These vesicles are then rapidly retrieved at the site of release, and are immediately available for re-release (33) At the ultrastructural level, many of the vesicles docked at the presynap-tic active zone tend to be smaller, as though they had just released some of their contents at the time of fixation (19)

Fig 3 Excitatory synapse revealed through ssTEM of a mushroom-shaped dendritic spine and

its corresponding presynaptic axonal bouton (a) A characteristic non-docked vesicle (arrow)

in the presynaptic axonal bouton (b) A docked vesicle (curved arrow) across from the postsynaptic density (PSD, chevron); on the spine head (arrow) (c) The synaptic cleft (wiggly

arrow) located between the plasma membranes of the spine head and presynaptic bouton

contains dense staining material, presumably composed of adhesion molecules and portions of

receptors (d) Serial section between (c) and (e) illustrates astroglial process (*) at the base of the spine head (e) Curved arrow illustrates another docked vesicle (f) Extracellular space

(arrow) does not contain the dense-staining material found in the synaptic cleft (g) Coated pit

(straight arrow) at a site of endocytosis on side of the bouton away from the active zone;

docked vesicle (curved arrow) (h) Gray surface of the plasma membrane (open arrow)

viewed en face where it caps the head of the dendritic spine In (a) the astroglial process

neurotransmitter; in all other sections the perisynaptic astroglial process is labeled with (*) (Adapted from (19))

2.3 Presynaptic Vesicles – Inhibitory Synapses

Fig 4 Inhibitory symmetric synapses in mature hippocampus with thin pre- and postsynaptic densities and pleiomorphic vesicles (a) Two inhibitory synapses on the pyramidal cell bodies

(open circles) (b) Symmetric synapse (open circle) with flattened vesicles (red arrows) and

asymmetric synapse (closed circle) with thicker PSD and larger rounder presynaptic vesicles

These synapses are located directly on the dendritic shaft of a nonspiny interneuron in mature hippocampal area CA1 (Modified from (16))

Inhibitory presynaptic boutons contain smaller, pleiomorphic vesicles having both round and flattened shapes in aldehyde-fixed tissue (Fig 4) The pleiomorphic vesicles usually contain the neurotransmitters GABA or glycine Inhibitory synapses are most abundant at the neuronal soma (Fig 4a) and proximal dendritic zones (see chapter by Arancibia-Carcamo, Triller and Kittler, this volume) In addition, inhibitory synapses can be interspersed among excitatory synapses (in the hippocampus about 1 inhibitory per 10–20 excitatory synapses) along a dendrite (Fig 4b) Occasionally, inhibitory synapses are located at the axonal hillock, where activation of one or a small number of inhibitory synapses can regulate neuronal cell firing at their axons In some brain re-gions an inhibitory synapse occurs on the necks of some dendritic spines, whether they veto excitatory activation likely depends on their frequency and the specific circuit involved (12, 24, 46) Neurosecretory peptides and some neurotransmitters are local-ized to the cytoplasm surrounding the pleiomorphic vesicles of inhibitory synapses, or

in large dense core vesicles (~100 nm) (10, 43) If axons use these large secretory DCVs then more of them occur in each axonal bouton

2.4 Presynaptic Small Dense Core Vesicles as Active Zone Transporters

Small dense core vesicles (~80 nm) are distinct from large DCVs both in size and frequency (Fig 5) Small DCVs are present in only about 20% of mature presynaptic axons, and when present, only 1–10 vesicles occur in a fully reconstructed axonal bouton The outer membranes of small DCVs label with antibodies to proteins lo-cated at the presynaptic active zone, such as piccolo and bassoon, and they are preva-lent along axons in the developing nervous system; hence the small DCVs are thought to be a local source of new presynaptic active zones (1, 50) (see chapter by Shen and Garner, this volume) Recent work has shown that there are fewer small DCVs during rapid synapse formation in the mature hippocampus in further support

of their role in local delivery of presynaptic active-zones (39)

Fig 5 Small dense core vesicles (arrowheads) at excitatory synapses in the apical dendritic field

of mature hippocampus (CA1) (a, b) Low and higher power views of a small dense core vesicle (dcv) in typical location near plasma membrane but away from active zone (c, d) Small DCV located near the beginning of an inter-varicosity region, suggesting it might be in transit (e, f)

Two small DCVs in a presynaptic axonal bouton One of these vesicles clearly illustrates small

‘spicules’ emanating from its surface (g, h) Small DCV located within one vesicle diameter of

an active zone DCVs are rarely located this close to an active zone, but show that they might also be involved in synapse enlargement, not just new synapse formation Scale bar in (g) is for (a, c, and e) Scale bar in (h) is also for (b, d, and f) (Adapted from (38))

2.5 Local Protein Synthesis in Presynaptic Boutons?

Although polyribosomes are not a prominent component of presynaptic axonal boutons in the central nervous system, mRNAs have been localized to them (32) In squid giant axons, local protein synthesis machinery appears to derive from the en-sheathing glia (9, 14) Detailed three-dimensional reconstructions will be required to learn whether isolated polyribosomes are directed into vertebrate presynaptic axonal boutons to allow for local protein synthesis, similar to that observed in dendritic spines (see below)

2.6 Nonsynaptic, Single Synaptic and Multisynaptic Axonal Boutons

Axonal boutons containing clear synaptic vesicles, mitochondria, dense core cles, and multivesicular bodies occur both with and without postsynaptic partners (Fig 6) In the mature hippocampus, about 96% of the vesicle-containing boutons have at least one postsynaptic partner Single-synapse boutons predominate compris-ing about 75% of all vesicle-containing axonal boutons About 21% of vesicle-containing boutons are multi-synaptic while about 4% are non-synaptic boutons in the mature hippocampus (36, 37) There are more multisynaptic boutons when rapid synaptogenesis occurs in the hippocampus, such as that which occurs during the estrus cycle (49) or following the preparation of mature hippocampal slices under ice-cold conditions (22, 23, 31) Multisynaptic dendritic spines, receiving input from more than one presynaptic axon occur relatively frequently during development and under conditions of synaptogenesis in the mature hippocampus The axonal segment

vesi-in Fig 6c was from a hippocampal slice that had been prepared under ice-cold section conditions, which induces new synapses These findings suggest that rapid synaptogenesis in the mature hippocampus does not require de novo formation of presynaptic axons It is not known whether the nonsynaptic boutons also constitute a source of available presynaptic boutons to accommodate rapid synaptogenesis, if they represent vesicle clusters in transit between presynaptic varicosities, or if thery are sites of recent synapse loss Similarly, presynaptic axonal boutons vary in struc-ture along the axons from other brain regions such as cerebellar cortex (Fig 7, (48)) Parallel fiber axons synapse with dendritic spines of Purkinje cell spiny branchlets (Fig 7a–c; see also Figs 8d, 11c and f below for further discussion of these postsy-naptic spines) The climbing fiber axons that synapse along the proximal dendrite of the Purkinje cells have much larger, more irregularly shaped boutons (Fig 7d) than parallel fiber axons

c

b

d

Fig 7 Reconstructed axons from rat cerebellar cortex (a–c) Parallel fiber axons

(d) Climbing fiber axon Axons (translucent light blue); PSDs (red); vesicles (dark blue);

mitochondria (beige) In all images, the locations of docked vesicles are superimposed on the

‘enface’ red PSD reconstructions to the right side of each axon The scale bar is 1 micron for all 4 reconstructions (Adapted from (48))

Fig 6 Axonal segments from mature hippocampus (CA1) (a) This segment is 7.8 µm long with

2 single-synapse boutons (SSB), 2 nonsynaptic boutons (NSB), and 1 multiple synapse bouton

(MSB) shared by 2 postsynaptic spines from different dendrites (b) This axonal segment is 5.7

µm long with 1 SSB containing 3 small DCVs and a mitochondrion It is surrounded by 3 NSBs

(c) This segment is 5.5 µm long with 1 SSB similar to that in (b) and another SSB in an unusual

position along the neck of a multi-synaptic dendritic spine (Axons – green, vesicles – yellow,

mitochondria – light blue, DCVs – dark blue, PSDs – red; Adapted from (38))

3 Postsynaptic Structure and Composition

3.1 Diversity in Postsynaptic Dendritic Spine Structure

Postsynaptic structure is highly diverse among synapses on the same dendrite and across different cell types in the brain For example, neighboring dendritic spines on

a single hippocampal dendritic segment can vary more than 10 fold in their sions (e.g Figs 2b, 11d) Large highly branched dendritic spines are the postsynaptic partners of mossy fibers in the hippocampal area CA3 (Fig 8a–c) The dendritic spines occurring along the spiny branchlets of cerebellar Purkinje cells vary widely

dimen-in their size, yet they are more uniformly club-shaped (Fig 8d) These cerebellar spines can be branched; however, the head of each branch on the spine is also club shaped (for more examples see (17)) Similarly, dendritic spines distributed along the same presynaptic axon vary greatly in their dimensions (Fig 6) The size of the den-dritic spine and its synapse correlates nearly perfectly with the number of vesicles in the presynaptic bouton for all brain regions tested so far (Fig 9) Thus, it is of great interest to understand the rules that govern local changes in synaptic structure and how they are coordinated between the pre- and postsynaptic structures One strategy has been to investigate the impact of altering the molecular composition of neurons

to determine the impact on dendritic spine structure (reviewed in (4, 42)) Another strategy has been to compare the composition of subcellular organelles among den-dritic spines and synapses of differing morphologies, during different stages of de-velopment, and during synaptic plasticity Our focus in this chapter is on this second strategy

Fig 8 3D reconstructions of diverse dendritic spines (a) Electron micrograph illustrating

CA3 pyramidal cell dendritic shaft (den), the origin of a complex branched spine (arrow head)

and mossy fiber (MF) bouton that synapses with spine heads from this and other dendrites

Scale = 1 µm (b) CA3 spine which contains microtubules (mt), polyribosomes (rib), and a

tubule of SER that becomes connected to a spine apparatus (sa) The fourth head connected to

this spine on a different serial section, has a multivesicular body (mvb) (c) CA3 spine (gold) with multiple PSDs (light blue) distributed across its many heads (d) Segment of spiney

branchlet from a cerebellar Purkinje cell (PSDs red, dendrite and spines beige) Scale in (b) is

1 µm for (b, c, and d) (a–c) are Adapted from (5); and (d) is Adapted from http://synapses.clm utexas.edu, courtesy J Spacek

3.2 Correlation Between PSD Area, Spine Head Volume,

and Vesicles in Presynaptic Axon

Fig 9 Strong correlation between PSD area, spine volume, and t number of presynaptic

vesi-cles This dataset is from hippocampal area CA1, but these relationships hold true across all brain regions evaluated so far, including synapses in hippocampal area CA3, cerebellar cortex, striatum, and neocortex (Adapted from (18, 26))

The postsynaptic density (PSD) contains a host of important receptors, scaffolding proteins, and signaling complexes that vary from synapse to synapse (21) The diver-sity in molecular composition from synapse to synapse is great, with hundreds of different proteins having been identified in the PSD In aldehyde-fixed tissue, excita-tory synapses are characterized by a thick PSD relative to the thinner presynaptic active zone, and hence are called ‘asymmetric synapses’ (Figs 1, 3, 5, 8, 10, 11, 12,

14, see also (15, 30)) Inhibitory synapses usually have equally thin PSDs and synaptic active zones and hence are referred to as ‘symmetric synapses’ (Fig 4) These relationships are highly diverse and the degree of pre- and postsynaptic thick-ening varies greatly even among excitatory or inhibitory synapses in the same brain region ((6), http://synapses.clm.utexas.edu/anatomy/chemical/colh.htm) Importantly, PSD thickness is also sensitive to experimental manipulations that are known to impact the abundance of PSD molecules, such as the calcium calmodulin dependent protein kinase type II (CaMKII), which are highly enriched in the PSD (28)

pre-The reconstructed surface area of the PSD correlates nearly perfectly with the volume and surface area of the spine head (Fig 9, (18, 26)) The PSD area and spine head volume also correlate nearly perfectly with the total number of presynaptic vesicles and the number of vesicles docked at the presynaptic active zone (18, 19, 26, 35) These observations suggest a strong structure-function relationship between dendritic spines and their presynaptic axons Exactly how this important relationship

is achieved and maintained during development and synaptic plasticity is not known

3.3 Polyribosomes – Local Protein Synthesis in Dendrites and Spines

Polyribosomes occur in some, but certainly not all dendritic spines (41) They are sites where translation could have been occurring at the time of tissue fixation Ri-bosomes are identified in dendrites and spines as 10–25 nm electron dense spheres surrounded by a gray halo; and polyribosomes are identified in ssTEM as having at least 3 ribosomes (Fig 10ab) Free polyribosomes synthesize cytoplasmic proteins, like those in the PSD and have been found to increase in frequency during plasticity such as LTP (27) Bound polyribosomes are associated with endoplasmic reticulum and synthesize integral proteins, such as receptors Occasionally polyribosomes have been detected in the vicinity of a spine apparatus (Fig 10b, 11a) Quantitative ssTEM is required to ascertain the relative frequencies of free and bound polyri-bosomes in spines and at other synapse types in the developing and mature brain and during plasticity such as learning and memory which requires local protein synthesis

Fig 10 Polyribosomes are sites of local protein synthesis present in some dendritic spines (a)

Head of a thin hippocampal dendritic spine with multiple polyribosomes (arrows; D, dendritic

shaft) (b) Polyribosome (arrow) located amidst a hippocampal spine apparatus (SA) (c) Only

one of the reconstructed spines on this dendrite had polyribosomes (ribosomes – black

spheres, PSDs – red) (Adapted from http://synapses.clm.utexas.edu/anatomy/ribosome/

ribo.stm courtesy J Spacek) Scale in (b) is about 0.5 µm for all three images

3.4 Diversity in Composition of Smooth Endoplasmic Reticulum

and Endosomal Compartments in Dendrites and Spines

Some dendritic spines contain smooth endoplasmic reticulum (SER), which tions as an internal calcium store In the hippocampus, only about 15% of all spines contain SER, and typically only the largest spines contain SER (40) In contrast, nearly all cerebellar dendritic spines contain SER (17) The spine apparatus is an enigmatic organelle variously thought to be involved in the regulation of calcium, synthesis of proteins, and post-translational modification of proteins like the Golgi apparatus (Fig 11a–d) It contains SER-like membranes arranged in laminae sepa-rated by dense-staining bars that contain the actin-binding protein synaptopodin which is necessary for maintaining its laminar structure and for positioning it into spines (13) Only large hippocampal and cortical spines contain a spine apparatus; in contrast the SER of cerebellar spines forms a reticular membranous network without the dense-staining laminations found in the spine apparatus (Fig 11ef) Total spine volume is well-correlated with the volume of SER, suggesting a functional role in maintaining ionic balance of cerebellar spines The presence or absence of SER might also account for the relative stability of cerebellar spines, and plasticity among hippocampal spines (3)

func-There are many other membrane-bound organelles present in dendrites and spines Many of these organelles are endosomes as revealed by tracking internalized gold particles conjugated with bovine serum albumin that were injected into the extracellular space of a PN21 hippocampal slice (7) The presence of gold particles

in the intracellular organelles unambiguously identifies them as endosomal or cling compartments inside dendrites and spines (Fig 12a–d) In contrast, the SER is

recy-a continuous membrrecy-ane-bound reticulum threcy-at is erecy-asily distinguished upon struction from the discrete tubules and vesicles of endosomal compartments (Fig 12e)

recon-Fig 11 Mushroom spines with spine apparatus (a) A spine apparatus is characterized by the

presence of SER laminated with dense staining material (thick filled arrow) At the base of this spine apparatus is rough endoplasmic reticulum (thin arrowhead) At the edge of the PSD

there are reciprocal coated structures The curved arrow indicates a double-walled coated vesicle, containing a small bit of the postsynaptic spine, or ‘spinule’, projecting from the spine

into the presynaptic axon The open short arrow indicates a coated pit, endocytosing into the

postsynaptic dendritic spine (b) Longitudinal section through a highly laminated spine

apparatus Straight arrows indicate SER; wiggly arrows indicate dense-staining material

(c) Reconstruction of a short hippocampal CA1 dendritic segment with 2 small thin spine and

1 large mushroom spine with surfaced PSDs (red) (d) Translucent surface of the dendrite and

spine reveals the SER (lavender) and mitochondrion (yellow) in the dendritic shaft Only the

large mushroom spine has SER, which forms a spine apparatus (pink) in this example (e) EM

image of SER in dendrite and spines of a spiney branchlet from a cerebellar Purkinje cell

Other spine origins emerging from this dendrite are indicated with thin arrows on the left (f)

Reconstruction of cerebellar spines and their reticulum of SER inside PSDs are red dendrite and spines are gray and ser is yellow in these spines ((a–d) adapted from (40); (e) is adapted

from (17); and (f) is from http://synapses.clm.utexas.edu/anatomy/compare/compare

Fig 12 Diversity among spines in composition of SER and endosomal compartments (a–d)

Large vesicles (lv), tubules with coated buds, amorphous vesicular clumps (avc) and multivesicular bodies with tubules all contained gold particles, endocytosed from the extracellular space Smooth endoplasmic reticulum (ser) never contained gold particles,

indicating that these two intracellular membranous compartments are not connected (e) Top,

reconstruction of SER in a dendrite and associated dendritic spines; about 14% of

hippocampal dendritic spines contained SER Bottom, reconstruction of a PN21 dendrite (gray) with endosomal structures (red) including a sorting complex (SC) which contains tubules, vesicles, a multivesicular body, and coated buds Green spheres represent the

locations of the smaller, smooth vesicles, presumably exocytic in function (Adapted from (7))

Scale in (a) is 0.5 µm for (a–d), and in (f) it is 1 um for (e) and (f)

Quantitative work shows that about 50 percent of normal hippocampal dendritic spines from perfusion fixed brain contain no membrane bound organelles; and a distinction among spines that contain the different types of organelles (Fig 13a) Interestingly, the amount of SER also varies along the shaft of the parent dendritic shaft (Fig 13b) Recent work shows endosomal compartments to be dynamically regulated during synaptic plasticity (29) It will be interesting to learn whether the presence or absence of SER is also dynamically regulated in spines during synaptic plasticity

b

Section number

a

Fig 13. Variation in the organization of SER and other membrane bound organelles in

hippocampal dendrites (a) Venn diagram showing the relative frequency of membrane-bound

organelles inside hippocampal dendritic spines About 49% of dendritic spines contain one or more membranous organelles It represents an average of all of the dendritic spines that were reconstructed from PN15, PN21, and Adult area CA1 of the rat hippocampus This

relationship is about the same at all three ages in perfusion fixed brain, in vivo (b) Relative

volume of SER in the dendritic shaft varies by type and number of dendritic spines along the length At the left is a schematic diagram of this dendrite with the location of thin spine origins along the dendritic segment illustrated to the left as small lollipops and mushroom spine locations illustrated by larger lollipops pointing to the right The graph illustrates the cross- sectional area of SER in each section along the length of the dendrite The three-dimensional

reconstruction of the SER in gray also illustrates its non-uniform distribution in the dendritic

shaft (Diagram in (a) is Modified from (6); and in (b) from (40))

4 Perisynaptic Astroglial Processes

In the cerebellum, nearly all synapses are enveloped by perisynaptic astroglial esses (Fig 14ab) Astroglial processes are identified by their light cytoplasm and the presence of dark glycogen granules and astrocytic fibrils in larger processes They end in thin or flattened processes that easily interdigitate among axons and dendrites throughout the brain In contrast, less than 50% of neocortical and hippocampal dendritic spine synapses have astroglial processes at their perimeters (Fig 14cd)

proc-Fig 14 Perisynaptic astroglial processes In cerebellar cortex, astroglial processes surround the synaptic cleft of nearly all synapses made by (a) climbing fibers or (b) parallel fibers In

contrast, only about half of all hippocampal synapses have astroglial processes at the cleft,

even between neighboring synapses (e.g s1Æ s2 Red arrow; s3 has no perisynaptic astroglia

at the presynaptic bouton, the postsynaptic spine, or the perisynaptic cleft region Scale bar = 1

um for all images) (a and b color schema: presynaptic axon (blue), postsynaptic spine (pink),

perisynaptic astroglia (yellow); (c) and (d) color scheme: presynaptic axon (green);

postsynaptic density (red); astroglia (blue)) (a, b adapted from (48)); (c, d adapted from (44))

Summary

In this chapter I have described the highly diverse structure and composition of synaptic axons, postsynaptic spines and dendrites, and perisynaptic astroglial proc-esses Postsynaptic size is proportional to presynaptic vesicle content and larger synaptic dendritic spines contain more of the subcellular organelles needed to re-model and sustain them Perisynaptic astroglial processes also appear to be key play-ers in determining synapse size Although synapse structure may appear static in electron micrographs, it is highly dynamic, undergoing dramatic changes in gross morphology, as well as, the positioning of subcellular organelles during basic synap-tic transmission and in response to conditions that elicit changes in synapse function

pre-or ‘synaptic plasticity’ Similarly, the structure and composition of synapses can vary across brain regions, and change dynamically during development, maturation, nor-mal aging, neurological disorders and trauma Careful consideration of this structural diversity and plasticity is leading to new understanding about synapse formation, growth, maintenance and elimination

This differential arrangement of perisynaptic astroglia suggests that some apses are highly regulated and supported by this important partner in the tripartite synapse, whereas other synapses are either more selective, unable to attract astroglial processes, or actively repel them (44) Exactly how this relationship is regulated by local synaptic plasticity remains to be determined, however, attraction to high concen-trations of extracellular glutamate is a likely candidate mechanism (8) In the mature hippocampus, those synapses that have a perisynaptic astroglial process at its perime-ter, are on average significantly larger than synapses without (47) Importantly, size is associated with the presence of an astroglial process juxtaposed to the postsynaptic spine and/or synaptic cleft; not the degree to which the astroglial process surrounds the synapse (Fig 14de) In fact, very large hippocampal synapses might have only a small fraction of their perimeters surrounded by an astroglial process (compare Fig 14de) These arrangements suggest that the perisynaptic astroglial processes do more than simply delineate boundaries between synapses to prevent spillover of neuro-transmitter and crosstalk and are likely crucial partners in sustaining synapses (2) One intriguing possibility, small, new, relatively unstable synapses in hippocampus or cortex, share the extracellular products released during synaptic transmission from other synapses, until they grow big enough to attract their own stabilizing astroglial processes (see chapter by Eroglu, Barres and Stevens, this volume)

syn-References

34:333–6

Curr Opin Neurobiol 2005; 15:542–8

Opin Neurobiol 2007; 17(3):381–6

4 Bourne JN, Harris KM Balancing structure and function at hippocampal dendritic

spines Annu Rev Neurosci 2008; 31:47–67

of CA3 branched dendritic spines and their synaptic relationships with mossy fiber

bou-tons in the rat hippocampus J Comp Neurol 1992; 325:169–82

visual cortex An electron microscope study Brain Res 1968; 9:268–87

hippo-campal dendritic spines from a widespread rather than a local store of recycling

mem-brane J Neurosci 2002; 22:2215–24

causes filopodia formation in cultured hippocampal astrocytes Glia 1990; 3:322–34

pre-synaptic endings of the synaptosomal fraction from squid brain J Neurosci 1997;

17:7694–702

10 Crivellato E, Nico B, Ribatti D Ultrastructural evidence of piecemeal degranulation in

large dense-core vesicles of brain neurons Anat Embryol (Berl) 2005; 210:25–34

11 De Camilli P, Slepnev VI, Shupliakov O et al The structure of synapses In: Cowen

WM, Sudhof TC, Stevens CF, eds The structure of synapses Baltimore, MD: The Johns

Hopkins University Press, 2000; 89–133

12 Dehay C, Douglas RJ, Martin KA et al Excitation by geniculocortical synapses is not

'vetoed' at the level of dendritic spines in cat visual cortex J Physiol 1991; 440:723–34

13 Deller T, Orth CB, Del TD et al A role for synaptopodin and the spine apparatus in

hippocampal synaptic plasticity Ann Anat 2007; 189:5–16

14 Eyman M, Cefaliello C, Ferrara E et al Local synthesis of axonal and presynaptic RNA

in squid model systems Eur J Neurosci 2007; 25:341–50

15 Gray EG Axo-somatic and axo-dendritic synapses of the cerebral cortex: An electron

microscopic study J Anat 1959; 93:420–33

16 Harris KM, Landis DM Membrane structure at synaptic junctions in area CA1 of the rat

hippocampus Neuroscience 1986; 19:857–72

17 Harris KM, Stevens JK Dendritic spines of rat cerebellar Purkinje cells: Serial electron

microscopy with reference to their biophysical characteristics J Neurosci 1988;

8:4455–69

18 Harris KM, Stevens JK Dendritic spines of CA1 pyramidal cells in the rat hippocampus:

serial electron microscopy with reference to their biophysical characteristics J

Neuro-sci 1989; 9:2982–97

19 Harris KM, Sultan P Variation in number, location, and size of synaptic vesicles vides an anatomical basis for the non-uniform probability of release at hippocampal CA1

pro-synapses J Neuropharmacol 1995; 34:1387–95

20 Heuser JE, Reese TS Structure of the synapse In: Kandel E, ed The handbook of

physi-ology, the Nervous system I American physiological society, 1977; 261–94

21 Kennedy MB Signal-processing machines at the postsynaptic density Science 2000;

290:750–4

22 Kirov SA, Petrak LJ, Fiala JC et al Dendritic spines disappear with chilling but

prolifer-ate excessively upon rewarming of mature hippocampus Neuroscience 2004; 127:

69–80

23 Kirov SA, Sorra KE, Harris KM Slices have more synapses than perfusion-fixed

hippo-campus from both young and mature rats J Neurosci 1999; 19:2876–86

24 Knott GW, Quairiaux C, Genoud C et al Formation of dendritic spines with GABAergic

synapses induced by whisker stimulation in adult mice Neuron 2002; 34:265–73

25 Landis DM, Hall AK, Weinstein LA et al The organization of cytoplasm at the

pre-synaptic active zone of a central nervous system synapse Neuron 1988; 1:201–9

26 Lisman J, Harris KM Quantal analysis and synaptic anatomy - integrating two views of

hippocampal plasticity Trends Neurosci 1993; 16:141–7

27 Ostroff LE, Fiala JC, Allwardt B et al Polyribosomes redistribute from dendritic shafts into spines with enlarged synapses during LTP in developing rat hippocampal slices

29 Park M, Salgado JM, Ostroff L et al Plasticity-induced growth of dendritic spines by

exocytic trafficking from recycling endosomes Neuron 2006; 52:817–30

30 Peters A, Palay SL, Webster HD The fine structure of the nervous system: The neurons

and supporting cells 3 ed Philadelphia, London, Toronto: W.B Saunders, Co., 1991; 1–

406

31 Petrak LJ, Harris KM, Kirov SA Synaptogenesis on mature hippocampal dendrites

occurs via filopodia and immature spines during blocked synaptic transmission J Comp

Neurol 2005; 484:183–90

32 Piper M, Holt C RNA translation in axons Ann Rev Cell Dev Biol 2004; 20:505–523

33 Rizzoli SO, Jahn R Kiss-and-run, Collapse and ‘Readily Retrievable’ Vesicles Traffic

2007; 8:1137–1144

34 Rostaing P, Real E, Siksou L et al Analysis of synaptic ultrastructure without fixative

using high-pressure freezing and tomography Eur J Neurosci 2006; 24:3463–74

35 Schikorski T, Stevens CF Morphological correlates of functionally defined synaptic

vesicle populations Nat Neurosci 2001; 4:391–5

36 Shepherd GM, Harris KM Three-dimensional structure and composition of CA3 >CA1 axons in rat hippocampal slices: implications for presynaptic connectivity and compart-

40 Spacek J, Harris KM Three-dimensional organization of smooth endoplasmic reticulum

in hippocampal CA1 dendrites and dendritic spines of the immature and mature rat J

Neurosci 1997; 17:190–203

41 Steward O, Levy WB Preferential localization of polyribosomes under the base of

den-dritic spines in granule cells of the dentate gyrus J Neurosci 1982; 2:284–91

42 Tada T, Sheng M Molecular mechanisms of dendritic spine morphogenesis Curr Opin

Neurobiol 2006; 16(1): 95–101

43 Torrealba F, Carrasco MA A review on electron microscopy and neurotransmitter

sys-tems Brain Res Brain Res Rev 2004; 47:5–17

44 Ventura R, Harris KM Three-dimensional relationships between hippocampal synapses

and astrocytes J Neurosci 1999; 19:6897–906

45 Volterra A, Magistretti PJ, Haydon PG The Tripartite Synapse Oxford: Oxford

Univer-sity Press, 2002

46 Wilson CJ, Groves PM, Kitai ST et al Three dimensional structure of dendritic spines in

rat striatum J Neurosci 1983; 3:383–98

47 Witcher MR, Kirov SA, Harris KM Plasticity of perisynaptic astroglia during

synapto-genesis in the mature rat hippocampus Glia 2007; 55:13–23

48 Xu-Friedman MA, Harris KM, Regehr WG Three-dimensional comparison of structural characteristics at depressing and facilitating synapses onto cerebellar Purkinje

ultra-cells J Neurosci 2001; 21:6666–72

49 Yankova M, Hart SA, Woolley CS Estrogen increases synaptic connectivity between single presynaptic inputs and multiple postsynaptic CA1 pyramidal cells: a serial elec-

tron – microscopic study Proc Natl Acad Sci U.S.A 2001; 98:3525–30

50 Ziv NE, Garner CC Cellular and molecular mechanisms of presynaptic assembly Nat

Rev Neurosci 2004; 5:385–99

The Role of Glutamate Transporters in Synaptic

Transmission

725 N Wolfe St., WBSB 1003, Baltimore, MD 21205, USA, dbergles@jhmi.edu

CA 94158-2517, USA, robert.edwards@ucsf.edu

1 Introduction

Exocytotic release of glutamate from preformed synaptic vesicles confers great speed on synaptic signaling, but also requires mechanisms to maintain the pool of available synaptic vesicles and remove glutamate from the extracellular space Since the axon terminal is in general located at a great distance from the cell body, synapses rely on the local recycling of both synaptic vesicle membrane and their neurotransmitter contents In this review, we address recycling of the excitatory transmitter glutamate, focusing on transport activities that operate at the plasma membrane and the synaptic vesicle, and the role that these transporters play in determining the amplitude and time course of synaptic responses

2 Glutamate Transport into Synaptic Vesicles and the Presynaptic Regulation of Quantal Size

Although analysis of the neuromuscular junction originally suggested that the response to release of a single vesicle filled with neurotransmitter is fixed, and hence represents the elemental “quantum” of synaptic transmission (109), considerable work has now shown that quantal size can change as a function of activity, contributing to such forms of plasticity as long-term potentiation (124) Nonetheless, the locus for this regulation is postsynaptic, and involves changes in receptor number

or sensitivity More recently, it has become clear that changes in quantal size can also reflect presynaptic changes in vesicle filling

It has long been appreciated that changes in the amount of neuromodulator released per vesicle can have profound consequences for the activation of receptors

at a distance from the release site Many G protein-coupled receptors have a relatively high affinity for their peptide and monoamine ligands, but the small

It is important to note that the mechanism of vesicular release poses several inherent problems Large amounts of transmitter per vesicle will result in the activation of more receptors, but high rates of firing will also deplete transmitter from the terminal unless it is actively replaced by, for example, recycling or biosynthesis At the same time, vesicular transport is generally slow, and may limit

amounts of ligand that actually impinge on receptors are not likely to saturate binding As a result, the release of more modulator activates more receptors, and considerable attention has focused on the regulation of quantal size for monoamines, taking advantage of electrochemical detection to measure dopamine release directly and in real time (181)

It has been less clear whether changes in vesicle filling with classical transmitters such as acetylcholine, GABA and glutamate make a difference in the postsynaptic response These transmitters are generally released in close apposition to postsynaptic receptors, many of which are ionotropic and have a high affinity for ligand (such as NMDA receptors for glutamate) If receptors are normally saturated

by the contents of a single vesicle, packaging more transmitter will have no effect on the postsynaptic response

If receptors are saturated, this will tend to reduce the variation in postsynaptic response However, quantal size exhibits considerable variation, particularly at central synapses Although this might result from variation in the distance of different synapses from the recording electrode, due to differences in electrotonic filtering, as well as variation in release probability and the number of receptors at each synapse, a number of observations have demonstrated that variation in quantal size is intrinsic to a single synapse Focal stimulation of one bouton, or localized dendritic recording, both show variation in quantal size similar to that observed from electrical stimulation of release from multiple boutons (17, 63, 120, 121) Increased cytosolic glutamate in the presynaptic terminal also increases quantal size at the calyx

iding additional evidence against receptor saturation Remarkably, a single vesicle filled with glutamate fails to saturate low-affinity AMPA receptors as well as high-affinity NMDA receptors (123, 132) Consistent with this, AMPA and NMDA responses are highly correlated at individual synapses, supporting a presynaptic locus for the variation GABA receptors at many (but not all) inhibitory synapses also appear not to be saturated by

a single vesicle (14, 67, 79)

How can synaptic release fail to saturate receptors? Although the concentration

of transmitter achieved in the synaptic cleft is high, the receptors are closely apposed

to the release site, and many are of high affinity, the peak concentration of transmitter is very brief, so that only a few receptors become activated Regardless of the precise explanation, changes in the amount of transmitter per vesicle are thus predicted to have a major influence on the postsynaptic response

The amount of neurotransmitter released from a synaptic vesicle may be

controlled either before or after the fusion event After fusion, premature closure of

the pore may interrupt the full release of vesicle contents Indeed, the exocytosis of large dense core vesicles frequently exhibits “kiss-and-run”, but this mechanism remains controversial for small synaptic vesicles, and the topic has recently been

reviewed elsewhere (60, 82) This review focuses on changes in quantal size before

fusion with the plasma membrane, that involve direct changes in vesicle filling

of Held in the auditory pathway (99), prov

refilling if vesicles recycle quickly, even at concentrations of cytosolic transmitter that saturate the transport mechanism Subsaturating cytosolic concentrations will further slow refilling and release However, low cytosolic concentrations may be important to prevent the oxidation and toxicity of monoamines such as dopamine (136), and this is compensated for by the ability of the vesicular monoamine

the lumen than the cytoplasm Other classical transmitters including glutamate produce toxicity through a specific interaction with cell surface receptors, and can therefore be tolerated at higher levels in the cytoplasm This reduces the magnitude

of the concentration gradient required to fill vesicles with glutamate, and presumably also speeds filling We will therefore consider now the factors that influence vesicle

gradient that drives transport, glutamate transport itself, and finally, the physiological regulation of these mechanisms and their role in synaptic plasticity The amount of transmitter achieved inside secretory vesicles indeed reflects the cytosolic concentration of transmitter, the driving force, transport into and non-specific leakage across the vesicle membrane

2.1 Cytosolic Glutamate: Biosynthesis and Recycling

Like any enzymatic reaction, the concentration of lumenal transmitter (product) depends directly on the cytosolic concentration (substrate) In the case of monoamines, electrochemistry has recently made it possible to measure cytosolic concentrations directly, by inserting a small carbon fiber electrode into a patch pipette (136) However, glutamate does not oxidize with the properties required for electrochemical detection, and less direct methods have therefore been used to estimate cytosolic concentrations Immunocytochemistry using specific antibodies to glutamate followed by electron microscopy originally suggested low millimolar levels in the cytoplasm (173, 179) More recently, dialysis of the calyx of Held with

50 mM glutamate increased the postsynaptic response of immature animals toward that observed by more mature, suggesting that cytosolic glutamate almost certainly exceeds 1 mM and might even approach 50 mM (99, 222)

2.1.1 Glutaminase

recycling In the case of glutamate, classical studies have shown that the preponderance of glutamate released as transmitter derives from glutamine (81) The neuronal enzyme glutaminase (also referred to as the phosphate-activated or kidney glutaminase, PAG) converts glutamine to glutamate before transport into synaptic vesicles (42, 114), and a liver isoform may also contribute (45, 113) Interestingly, both isoforms can be regulated by inorganic phosphate (45), which converts the inactive monomer to active tetramer However, the loss of kidney glutaminase has shown remarkably little defect in excitatory transmission (129) The animals do not survive past birth, but the enzyme also has roles in nitrogen metabolism and pH regulation that may account for the observed lethality Indeed, excitatory