UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trafficking of scaffolding and adhesion proteins : the role of pre-assembled complexes and lateral diffusion… Gerrow, Kimberly A. 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2008_fall_gerrow_kimberly.pdf [ 7.27MB ]
Metadata
JSON: 24-1.0066686.json
JSON-LD: 24-1.0066686-ld.json
RDF/XML (Pretty): 24-1.0066686-rdf.xml
RDF/JSON: 24-1.0066686-rdf.json
Turtle: 24-1.0066686-turtle.txt
N-Triples: 24-1.0066686-rdf-ntriples.txt
Original Record: 24-1.0066686-source.json
Full Text
24-1.0066686-fulltext.txt
Citation
24-1.0066686.ris

Full Text

    Trafficking of scaffolding and adhesion proteins: The role of pre-assembled complexes and lateral diffusion during synapse development     by  Kimberly A. Gerrow  B.Sc., Queen¶s University, 1999     A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES  (Neuroscience)    THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  October 2008           © Kimberly A. Gerrow, 2008  ii   ABSTRACT  Part One: Role of a pre-formed scaffolding complex in excitatory synapse formation. In order to determine the role of non-synaptic clusters of postsynaptic proteins we monitored the trafficking of several candidate proteins implicated in synaptogenesis, when non-synaptic clusters of scaffold proteins are most abundant. We found a protein complex consisting of two populations that differ in their content, mobility, and involvement in synapse formation. One subpopulation is mobile and relies on actin transport for delivery to nascent and existing synapses. These mobile clusters contain the scaffolding proteins PSD-95, GKAP, and Shank. The second group consists of stationary non-synaptic scaffold complexes that mainly contain neuroligin-1, can recruit synaptophysin-containing axonal transport vesicles, and are readily transformed to functional presynaptic contacts that recycle the vital dye FM 4-64. These results postulate a mechanism whereby preformed scaffold protein complexes serve as predetermined postsynaptic hotspots for establishment of new functional excitatory synapses.  Part Two: Neuroligin trafficking in live neurons. The mechanisms that govern the differential trafficking and retention of neuroligin-1 to glutamatergic synapses and neuroligin-2 to GABAergic synapses remain unclear. In order to monitor the recruitment/retention of neuroligin-1 and -2 to synaptic sites, a site- specific biotinylation-based approach was utilized that allows for the visualization of surface proteins in live neurons with monovalent streptavidin. To quantify these changes, FRAP (fluorescence recovery after photo beaching) showed similar recovery rates for GFP-tagged neuroligins (representative of the total pool) compared to AP-tagged neuroligins (representative of the surface pool). The mobile pool of neuroligin-1 clusters was significantly larger than neuroligin-2 clusters and was depressed in older neurons. The mobility of neuroligin-1 clusters was influenced by the expression of specific scaffolding proteins. Exogenous expression of PSD-95 or S-SCAM, but not Shank, reduced the mobile faction of neuroligin-1, whereas the mobile fraction of neuroligin-2 was reduced by exogenous expression of S-SCAM. Interruption of actin and microtubule cytoskeleton interactions also decreased the recovery of both neuroligin-1 and -2. Thus the differential mobility of neuroligin molecules as well as association with particular scaffolding proteins and the cytoskeleton may contribute to their lateral diffusion and retention at particular synaptic sites.  iii TABLE OF CONTENTS  ABSTRACT ......................................................................................... ii TABLE OF CONTENTS..................................................................... iii LIST OF TABLES.............................................................................. vii LIST OF FIGURES........................................................................... viii LIST OF ABBREVIATIONS................................................................ x ACKNOWLEDGEMENTS ................................................................. xii CO-AUTHORSHIP STATEMENT.................................................... xiii  1 Introduction.................................................................................. 1 1.1 OVERVEIW OF THE SYNAPSE................................ ......................... 1 1.1.1 Synapse structure ................................ ................................ .............1 1.1.2 Steps of synapse formation................................ ............................... 3 1.2 DEVELOPMENT OF THE NEUROMUSCULAR JUNCTION..............6 1.2.1 Pre-patterning ................................ ................................ ...................7 1.2.2 Postsynaptic maturation and reciprocation................................ ........8 1.2.3 Synaptic refinement and re-modeling................................ ................8 1.3 DEVELOPMENT OF SYNAPSES IN THE CNS................................ ..9 1.3.1 Contact initiation................................ ................................ ................9 1.3.2 Recruitment of presynaptic and postsynaptic proteins ....................11 1.3.3 Maturation ................................ ................................ ....................... 16 1.3.4 Elimination................................ ................................ ....................... 17 1.4 PSD SCAFFOLDING PROTEINS................................ .....................19 1.4.1 Protein interactions of PSD-95 ................................ ........................ 20 1.4.2 Other PDZ proteins ................................ ................................ .........23 1.4.3 Role in synapse formation................................ ............................... 26 1.5 NEUROLIGINS AND NEUREXINS................................ ...................27 1.5.1 Structure of neuroligins ................................ ................................ ...27 1.5.2 Structure of neurexins ................................ ................................ .....29 1.5.3 Crystal structure and binding................................ ........................... 30 1.5.4 Spatial and temporal distribution ................................ .....................32 1.5.5 Synaptic function................................ ................................ .............34  1.5.6 In vivo function: Knock-out strategies................................ ................36 1.6 INHIBITORY SYNAPSE FORMATION ................................ .............37 1.6.1 Receptors................................ ................................ ........................ 38 1.6.2 Scaffolding proteins................................ ................................ .........39 1.6.3 Cell adhesion molecules ................................ ................................ .41 1.6.4 Presynaptic mechanisms ................................ ................................ 44 1.7 SYNAPTIC IMBALANCE IN AUTISM ................................ ...............45 1.7.1 Definition and etiology ................................ ................................ .....45 1.7.2 Genetic links................................ ................................ ....................45 1.7.3 Mouse models................................ ................................ .................47 1.8 RATIONALE AND HYPOTHESIS ................................ .....................48 1.9 REFERENCES ................................ ................................ .................49  iv 2 A preformed complex of postsynaptic scaffolding proteins is involved in glutamatergic synapse development ........................ 68 2.1 INTRODUCTION................................ ................................ ...............69 2.2 MATERIALS AND METHODS ................................ .......................... 72 2.2.1 cDNA constructs................................. ................................ .............72 2.2.2 Neuronal cell culture and transfection. ................................ ............72 2.2.3 Immunocytochemistry ................................ ................................ .....73 2.2.4 Assessment of presynaptic terminal function using FM 4-64. .........73 2.2.5 Imaging and analysis................................ ................................ .......74 2.3 RESULTS ................................ ................................ ......................... 75 2.3.1 PSD-95, GKAP, and Shank co-localize at non-synaptic sites in young neurons ................................ ................................ ................75 2.3.2 Mobile and stationary clusters containing PSD-95, GKAP and Shank  «««««««««««««««««««««««««««««.77 2.3.3 Stationary non-synaptic preformed scaffold complexes participate in development of functional presynaptic terminals............................. 79 2.3.4 Delivery of mobile synaptophysin clusters to contact sites apposed to stationary preformed scaffold complexes ................................ ....81 2.3.5 Stationary non-synaptic preformed scaffold complexes contain neuroligin-1 ................................ ................................ .....................86 2.3.6 Mobility of preformed scaffold clusters is actin dependent ..............90 2.3.7 Mobile preformed scaffold clusters are recruited to nascent and existing postsynaptic sites................................ ............................... 92 2.3.8 Knock Down of PSD-95 results in a decrease in GKAP, Shank and VGLUT clusters, and an increase in VGAT clusters positive for neuroligin-1 ................................ ................................ .....................94 2.4 DISCUSSION................................ ................................ ....................96 2.5 REFERENCES ................................ ................................ ...............103  3 Neuroligin trafficking on the surface: The role of diffusion in the trafficking of adhesion proteins. ........................................... 106 3.1 INTRODUCTION................................ ................................ .............107 3.2 MATERIALS AND METHODS ................................ ........................ 110 3.2.1 cDNA constructs................................ ................................ ............110 3.2.2 Neuronal cell culture and transfection ................................ ...........111 3.2.3 Immunocytochemistry and analysis................................ ...............112 3.2.4 Surface labeling with monovalent streptavidin .............................. 112 3.2.5 Fluorescence recovery after photo-beaching (FRAP) ...................115 3.2.6 Time-lapse ................................ ................................ ....................115 3.3 RESULTS ................................ ................................ ....................... 116 3.3.1 Neuroligin mobility is isoform specific and developmentally regulated  ««««««««««««««««««««««««««««..116 3.3.2 Surface neuroligin accounts for most of the mobile pool. ..............118 3.3.3 Recovery of scaffolding proteins and neuroligins are correlated. ..121 3.3.4 Neuroligin recovery at sites where scaffold proteins are diffuse....122  v 3.3.5 Neuroligins membrane mobility is determined by intracellular domains................................ ................................ ......................... 124 3.3.6 The mobile pool of neuroligin-1 is dependent on PDZ interactions.  ««««««««««««««««««««««««««««..126 3.3.7 Specific scaffold proteins influence neuroligin surface mobility .....128 3.3.8 Neuroligin mobility is differentially influenced by the actin and microtubule cytoskeleton................................. .............................. 132 3.3.9 Neuroligin-2 surface retention is determined by intracellular domains  ««««««««««««««««««««««««««««..135 3.4 DISCUSSION................................ ................................ ..................137 3.5 REFERENCES ................................ ................................ ...............144  4 Overall Discussion .................................................................. 149 4.1 SUMMARY OF FINDINGS................................ .............................. 150 4.2 ROLE OF PRE-ASSEMBLED SCAFFOLDING COMPLEXES IN SYNAPSE FORMATION................................ ................................ .152 4.2.1 How do pre-assembled complexes fit into out current knowledge of glutamatergic synapse formation?................................ .................152 4.2.2 What about other synapse types?................................ .................155 4.2.3 The connection between synapse formation and neuron morphology  ««««««««««««««««««««««««««««..157 4.3 ROLE OF SURFACE MOBILITY AND INTRACELLULAR INTERACTIONS FOR ADHESION PROTEIN LOCALIZATION .....160 4.3.1 How does this fit in with our current knowledge of adhesion protein trafficking?................................ ................................ .....................160 4.3.2 How does this fit in with our current knowledge of factors that affect neuroligin trafficking? ................................ ................................ ....161 4.3.3 Cytoplasmic interactions ................................ ............................... 164 4.4 THERAPEUTIC TARGET ................................ ............................... 166 4.5 CONCLUSION ................................ ................................ ................167 4.6 REFERENCES ................................ ................................ ...............168  5 Additional materials and methods ........................................ 173 5.1 FILTER SETS ................................ ................................ .................173 5.2 IMAGE ANALYSIS................................ ................................ ..........174 5.3 ANALYSIS OF EXPRESSION OF TAGGED PROTEINS ...............175 5.4 EVALUATION OF FM 4-64 ................................ ............................. 181 5.5 SYNTHESIS OF MONOVALENT STREPTADVIDIN ...................... 182 5.6 CYTOSKELETON DEPOLYMERIZATION ................................ .....184 5.7 REFERENCES ................................ ................................ ...............185  6 Other contributions................................................................. 186 6.1 D1 RECEPTOR ACTIVATION INCREASES AMPA RECEPTOR SURFACE EXPRESSION AND SYNAPTIC RETENTION..............186  vi 6.2 Als2-DEFICIENT MICE EXHIBIT DISTURBANCES IN ENDOSOMAL TRAFFICKING ASSOCIATED WITH MOTOR BEHAVIOURAL ABNORMALITIES. ................................ ................................ ..........189 6.3 A BALANCE BETWEEN EXCITATROY AND INHIBITORY SYNAPSES IS CONTROLLED BY PSD-95 AND NEUROLIGIN....192 6.4 NEUROLIGINS MEDIATE EXCITATROY AND INHIBITORY SYNAPSE FORMATION................................ ................................ .194 6.5 PRESYNAPTIC TRAFFICKING OF SYNAPTOTAGMIN I ..............194 6.6 REFERENCES ................................ ................................ ...............195   vii LIST OF TABLES  Table 1.1: Other synaptic PDZ proteins ................................ ................................ ......... 25 Table 1.2: Chromosome abnormalities associated with autism................................ ....... 47 Table 3.1: NL1 clustering by streptavidin. ................................ ................................ ... 113 Table 3.2: Effect of streptavidin on VGLUT1 clustering................................. ............. 114 Table 4.1: Synapse selectivity of neurexins and neuroligins................................. ........ 162 Table 4.2: Interactions between the cytoplasmic domains of neuroligins and various PDZ domains................................. ................................ ................................ ...................... 164   viii LIST OF FIGURES  Figure 1.1: Schematic diagram of a prototypical synapse................................. ................ 2 Figure 1.2: Stages in the formation of a glutamatergic synapse in the CNS. ..................... 4 Figure 1.3: Molecular constituents at the NMJ................................. ................................ 6 Figure 1.4: Initiation of synaptogenic contacts by axons and dendrites........................... 10 Figure 1.5:  Multiple mechanisms for the recruitment and stabilization of synaptic proteins to new sites of contact. ................................ ................................ ..................... 12 Figure 1.6: Different modes that regulate clustering of AMPA receptors at the synapse. 15 Figure 1.7: Dendritic spine morphology and signaling. ................................ .................. 16 Figure 1.8: Proteins from postsynaptic density (PSD) fractions................................. ..... 19 Figure 1.9: Schematic diagram of PDZ proteins at a mammalian excitatory synapse...... 21 Figure 1.10: Domain structure of neuroligins................................. ................................ 28 Figure 1.11: Neurexin domain structure................................. ................................ ........ 30 Figure 1.12: Structure of neuroligin-1................................. ................................ ........... 31 Figure 1.13: Structure of the NL1/Nrx1ȕ Complex. ................................ ....................... 32 Figure 1.14: Endogenous localization of neuroligin-1 and -2 in cultured hippocampal neurons................................. ................................ ................................ ......................... 33 Figure 1.15: Neuroligin-1 induces presynaptic contact formation................................. ..36 Figure 1.16: Postsynaptic gephyrin and its binding partners................................. .......... 40 Figure 2.1: Detection of non-synaptic clusters of postsynaptic scaffolding proteins. ...... 76 Figure 2.2: Quantification of the mobility of PSD-95, GKAP and Shank. ...................... 77 Figure 2.3: Visualization of mobile and stationary clusters containing PSD-95, GKAP and Shank................................. ................................ ................................ ..................... 78 Figure 2.4: Co-localization of PSD-95 and VGlut increases with developmental age. .... 80 Figure 2.5: Sites apposed to stationary non-synaptic scaffold clusters are readily transformed to active presynaptic terminals. ................................ ................................ ..81 Figure 2.6: Characterization of transport packets positive for synaptophysin DsRED..... 82 Figure 2.7: Time-lapse of SYN DsRED and PSD-95 GFP. ................................ ............ 84 Figure 2.8: Dynamics of a growth cone contacting PSD-95 GFP Clusters...................... 85 Figure 2.9: Neuroligin-1 is associated with Synaptic and non-synaptic preformed scaffold clusters. ................................ ................................ ................................ ......................... 86 Figure 2.10: Stationary clusters contain neuroligin-1 ................................ ..................... 87 Figure 2.11: Stationary scaffold clusters contain neuroligin-1. ................................ ....... 89 Figure 2.12: Stationary scaffold clusters containing neuroligin-1 recruits synaptophysin positive transport packets. ................................ ................................ ............................. 90 Figure 2.13: PSD-95 Movement is Actin Dependent................................. ..................... 91 Figure 2.14: Mobile preformed scaffold clusters are recruited to nascent and existing sites. ................................ ................................ ................................ .............................. 93 Figure 2.15: Knock-down of PSD-95 reduces clustering of GKAP, Shank, and VGLUT. ................................ ................................ ................................ ................................ ...... 95 Figure 2.16: A model summarizes the role of stationary and mobile preformed scaffold clusters in excitatory synapse development................................. ................................ .102 Figure 3.1: Monovalent streptavidin reduces aggregation of NL1. ............................... 113 Figure 3.2: Aggregation of NL1 by wild-type streptavidin reduces NL1 enhancement of presynaptic terminals................................. ................................ ................................ .. 114  ix Figure 3.3: Neuroligin mobility is isoform specific. ................................ ..................... 117 Figure 3.4: Neuroligin mobility is developmentally regulated................................. ..... 118 Figure 3.5: Surface neuroligin accounts for most of the mobile pool............................ 119 Figure 3.6: Neuroligin mobility on the surface is developmentally regulated. .............. 120 Figure 3.7: Correlation of NL recovery at the correct scaffold. ................................ .... 121 Figure 3.8: Neuroligin FRAP recovery at sites not associated with diffuse scaffold proteins, versus clustered sites with scaffolding proteins................................. ............. 123 Figure 3.9: Schematic of AP-tagged NL c-terminal deletions................................. ...... 124 Figure 3.10: Time-lapse of surface labeled NL1 and NL2 demonstrate different cluster dynamics. ................................ ................................ ................................ .................... 126 Figure 3.11: PDZ interactions influence the NL1 mobile fraction. ............................... 127 Figure 3.12: PSD-95 influences NL1 mobile pool in a PDZ-dependent manner. .......... 129 Figure 3.13: S-SCAM influences the mobile fraction of NL1 and NL2. ....................... 130 Figure 3.14: S-SCAM enhances clustering and function of neuroligins........................ 131 Figure 3.15: Neuroligin surface mobility is differentially influenced by the actin and microtubule cytoskeleton................................. ................................ ............................ 133 Figure 3.16: Scaffolding proteins are differentially regulated by the actin and microtubule cytoskeleton. ................................ ................................ ................................ ............... 134 Figure 3.17: Insertion of AP-tagged neuroligin-2................................. ........................ 135 Figure 3.18: Internalization of neuroligin-2 is dependent on c-terminal domains.......... 136 Figure 3.19: Schematic of interactions that influence the lateral mobility of neuroligins. ................................ ................................ ................................ ................................ .... 139 Figure 4.1: Axon filopodia preferentially extend toward and contact pre-patterned AChR clusters. ................................ ................................ ................................ ....................... 156 Figure 4.2:Dendritic arbor growth and synapse maturation are concurrent. .................. 158 Figure 5.1: Sample images for bleed-through and cross-excitation testing. .................. 173 Figure 5.2: Characterization of Expression of PSD-95 GFP, GKAP DsRED, and Shank CFP. ................................ ................................ ................................ ............................ 176 Figure 5.3: Characterization of Exogenous Expression of PSD-95 GFP, and neuroligin CFP. ................................ ................................ ................................ ............................ 177 Figure 5.4: Expression of DsRED synaptophysin................................. ........................ 178 Figure 5.5: AP-tagged NL expression analysis................................. ............................ 179 Figure 5.6: AP-tagged NL distribution................................. ................................ ........ 179 Figure 5.7: AP-tagged NL FRAP with GFP. ................................ ................................ 180 Figure 5.8: Fidelity of FM 4-64. ................................ ................................ .................. 181 Figure 5.9: Generation of monovalent streptavidin. ................................ ..................... 183 Figure 5.10: Effect of CytoB and NOCOD treatment on the overall morphology of neurons................................. ................................ ................................ ....................... 184 Figure 6.1: S845 in AP-GluR1 is important for the insertion and incorporation of AMPA receptors selectively at synaptic sites upon activation of D1 receptors. ........................ 188 Figure 6.2: Neurons from Als2±/± mice show disturbances of BDNF receptor (TrkB) endocytosis................................. ................................ ................................ ................. 190 Figure 6.3: Neurons from Als2±/± mice show disturbances of IGF1R endocytosis......... 191 Figure 6.4: Altered expression of PSD-95 influences the ratio of excitatory-to-inhibitory presynaptic contacts. ................................ ................................ ................................ ... 193   x LIST OF ABBREVIATIONS  AChR ±AcetylCholine Receptor AMPA - Alpha-amino-3-hydroxy-5-Methyl-4-isoxazolePropionic Acid AKAP - A-Kinase-Anchoring Protein AP ± Acceptor Peptide APB - AMPA receptor±Binding Protein ASD - Autism Spectrum Disorder BirA - Bacterially derived biotin ligase CAM ± Cell Adhesion Molecule CFP ± Cyan Fluorescent Protein CLAM - Cholinesterase-Like Adhesion Molecules CLD - Cholinesterase-Like Domain CNS ± Central Nervous System CytoB - Cytochalasin B DIV ± Day In Vito Dlc - Dynein light chain DLG ± Discs Large DsRED - Discosoma sp. red fluorescent protein EGF (Epidermal Growth Factor E/I ± Excitatory/Iinhibitory EM ± Electron Microscopy EZ - Endocytic Zone FRAP ± Fluorescence Recovery After Photobleaching GABA - γ-Aminobutyric Acid GABARAP - GABAA  -Receptor-Associated Protein GAP - GTPase Activating Protein GFP ± Green Fluorescent Protein GK ± Guanylate Kinase GKAP - Guanylate Kinase Associated Protein GRIP - Glutamate Receptor Interacting Protein KO ± Knock-Out  xi LNS - Laminin, Neurexin, Sex-hormone-binding protein LTD ± Long-Term Depression LTP ± Long-Term Potentiation MAGUK - Membrane-Associated Guanylate Kinase mSA ± monovalent Streptavidin MuSK ± Muscle Specific Kinase NL - Neuroligin NMDAR - N-methyl-D-aspartic acid Receptor NMJ ± Neuromuscular Junction NOCOD - Nocodozole NRX - Neurexin PDZ- PSD-95, Dlg and ZO-1 PICK - Protein kinase C±interacting protein PKA - cAMP-dependent protein kinase PSD ± Postsynaptic Density PTV ± Piccolo-Transport Vesicles SALM - Synaptic Adhesion-Like Molecules SAP - Synapse-Associated Protein siRNA ± small interfering RNA S-SCAM - Synaptic Scaffolding Molecule SNARE - Soluble NSF Attachment Receptor STV ± Synaptic Transport Vesicles SYN ± Synaptophysin SynCAM ± Synaptic Cell Adhesion Molecule TARP - Transmembrane AMPA Receptor±binding Protein TM - Transmembrane TTX - Tetrodotoxin VGAT ± Vesicular GABA Transporter VGluT - Vesicular Glutamate Transporter YFP ± Yellow Fluorescent Protein  xii ACKNOWLEDGEMENTS   Many thanks go to my supervisor Dr. Alaa El-Husseini, for his excellent supervision and support. I will greatly miss his mentorship and friendship.  I would also like to thank Dr. Tim Murphy for helping me in these final steps towards my thesis, as well as the other members of my advisory committee, Drs. Tim O¶Connor and Yutian Wang.  Of my many helpful fellow students, I want to give special thanks to Kun Huang, Marie- France Lise, Joshua Levinson, Rochelle and Dustin Hines, Catherine Campbell and Pamela Arstikaitis. We shared a wonderful lab, full of good advice, helping hands, and lasting friendship. I would also like to thank colleagues from the Kinsmen tower of power, including Oliver Prange, Andy Shih, Herman Fernandas, Austin Milnerwood, for their insightful help and procrastination skills. Furthermore, I want to thank Esther Yu for her assistance in neuronal cell culture preparation, and Rujun Kang for keeping our gaggle of students organized.  Lastly, I deeply want to thank my husband Dave for being so supportive in my academic pursuits, as well as my friends and family for their continuing encouragements.  During my dissertation I was supported by grants from the Michael Smith Foundation of Health Research (MSFHR) and Canadian Institute of Health Research (CIHR). I would like to thank these agencies for their financial support.  xiii CO-AUTHORSHIP STATEMENT   Chapter 2 The experiments in this chapter were conceived and designed by me. I analyzed the data, prepared the figures and wrote the manuscript. I developed all of the techniques, reagents, and conducted all of the experiments with the exceptions noted below. Preparation of neuronal cultures was assisted by lab technician, Esther Yu. Construction of GKAP constructs was by S. Nabi. siRNA transfection and neuroligin staining and analysis was performed by S. Romorini in collaboration with the Sala lab. Growth cone time-lapse was performed and analyzed by M. Colicos.   Chapter 3 The experiments in this chapter were conceived and designed by me. I analyzed the data, prepared the figures and wrote the manuscript. I developed all of the techniques, reagents, and conducted all of the experiments with the exceptions noted below. Preparation of neuronal cultures was assisted by lab technician, Esther Yu. Synthesis of monovalent streptavidin was done by M. Howarth, in collaboration with the lab of A.Y. Ting. Construction of neuroligin-2 c-terminal deletions was performed by J. Levinson, and later re-subcloned by myself and sequenced by H. Takahashi. The following FRAP experiments were performed by H.Takahashi and analyzed by H.Takahashi and myself: NL at no scaffold, PDZ deletion mutants, S-SCAM over expression, nocodozole and cytochalasinB.  1 1 Introduction  1.1 OVERVIEW OF THE SYNAPSE  Synapses are specialized intercellular junctions which provide the structural and functional basis for the formation and maintenance of the complex neural network in the brain. The number, location, and type of synapses formed are well controlled, since synaptic circuits are formed in a highly reproducible way, implying the existence of cellular and molecular mechanisms that determine this connectivity. Additionally, neurons develop and maintain distinct molecular compositions for each individual synapse, despite being often only microns apart (Aoki et al., 2001; Craig and Boudin, 2001; Craig et al., 2006; Friedman et al., 2000; Kim and Sheng, 2004; Nimchinsky et al., 2002; O'Brien et al., 1998; Ziv, 2001). Understanding the mechanisms that govern synapse formation, maturation, and maintenance are fundamental to our understanding of circuit formation and ultimately, brain function. Recent work has begun to unravel the mechanisms controlling formation of synapses, demonstrating the existence of multiple molecules that influence the many aspects of synapse formation, specificity and stability.  1.1.1 Synapse structure Synapses are asymmetric in their structure and function, making them distinct from other cell-cell junctions (Figure 1.1).  On one side is the axon or presynaptic terminal, which is engorged with synaptic vesicles containing neurotransmitters, the chemical basis of neuron to neuron communication. Several of these vesicles are docked on a specialized area of the presynaptic membrane called the active zone, attached by a complex matrix of proteins, primed for release. The release of a neurotransmitter is triggered by arrival of an action potential which produces an influx of calcium ions through voltage-dependent calcium-selective ion channels. Calcium ions then trigger a biochemical cascade which results in vesicle fusion with the presynaptic membrane, driven by the action of SNARE proteins, and releasing their contents to the synaptic cleft (Rosenmund et al., 2003;  2 Schneggenburger and Neher, 2005; Stevens, 2003). Released neurotransmitters diffuse across the small gap of the synaptic cleft to the cell membrane on the other side. At the other side, also known as the postsynaptic side, neurotransmitter receptors and a wide array of transmembrane, cytoskeletal and signaling proteins are clustered and poised to respond to the neurotransmitter released from the presynaptic terminal (Boeckers, 2006; Kennedy, 1997; Kornau et al., 1997; Okabe, 2007). The postsynaptic site can be found on the neuronal soma, dendrites, or initial segment of an axon, but can also be on a gland or muscle cell.    Figure 1.1: Schematic diagram of a prototypical synapse. Vesicles containing neurotransmitter sit docked at the active zone. Release of neurotransmitter is triggered by arrival of an action potential (1), which produces an influx of calcium ions through voltage-dependent, calcium-selective ion channels (2,3). Calcium then triggers a biochemical cascade which results in vesicle fusion with the presynaptic membrane, driven by the action of SNAREs (inset), and release their contents to the synaptic cleft (4). Receptors on the opposite side of the synaptic gap bind neurotransmitter molecules and respond by opening nearby ion channels in the post-synaptic cell membrane, causing ions to rush in or out and changing the local transmembrane potential of the cell (5). Whether a synapse is excitatory or  3 inhibitory depends on what type(s) of ion channel conduct the post-synaptic current display(s), which in turn is a function of the type of receptors and neurotransmitter employed at the synapse. Termination of the signal is mediated by uptake or degradation of the neurotransmitter from the synaptic cleft (6). Reproduced with permission from the artist, Astrid Elizabeth Vincent Andersen.    The transmitter released presynaptically must match the types of receptors and signaling molecules postsynaptically. Therefore, a form of trans-synaptic communication must exist for the presynaptic axon terminal to recognize its appropriate postsynaptic partner, and vice versa. Because neurons each receive millions of synapses, they are confronted with the major task of sorting and targeting pre- and postsynaptic proteins to appropriate sites. This precise, synapse-specific matching implicates that sorting and targeting mechanisms exist for the molecular constituents of different types of synapses, to ensure correct formation of neuronal circuits in the brain (Lardi-Studler and Fritschy, 2007).   1.1.2 Steps of synapse formation  Synapse formation is thought to involve several characteristic steps: contact initiation, recruitment of presynaptic and postsynaptic proteins, and stabilization (Figure 1.2) (Ahmari and Smith, 2002; Brose, 1999; Cline, 2005; Cline, 2001; Ferreira and Paganoni, 2002; Garner et al., 2002; Gerrow and El-Husseini, 2006; Goda, 2002; Levinson and El- Husseini, 2005b; Montgomery et al., 2004; Patterson, 2002; Sanes and Lichtman, 1999; Waites et al., 2005b; Zhang and Benson, 2000; Ziv, 2001).  First, an initial contact is made between an axon and the target postsynaptic cell. This has traditionally been described in terms of an axonal growth cone that sends out filopodia searching for a prospective partner. This view is a carry over from studies of the NMJ where the target (muscle) is relatively stationary (Sanes and Lichtman, 1999). Dendrites however, also extend growth cones and are lined with filopodia along their entire length. The dynamic protrusive behavior of these dendritic filopodia is consistent with an active ability to initiate contact formation (Fiala et al., 1998; Ziv and Smith, 1996). Regardless  4 of whether dendrites or axons are responsible for initiation of contact, target recognition is thought to be specific, since correct connectivity is essential to the function of a neural network.    Figure 1.2: Stages in the formation of a glutamatergic synapse in the CNS. (1) Axons often travel long and arduous pathways before reaching their target field. Dendrites also grow into complex arborizations. (2) Both axons and dendrites extend long finger-like protrusions called filopodia which are believed to be actively searching for appropriate targets.  (3) After appropriate presynaptic and postsynaptic targets make contact, proteins important for a functional synapse are recruited. On the presynaptic side, vesicles containing neurotransmitter and the accessory proteins mediating their fusion are assembled.  Directly apposed on the postsynaptic side, receptors and scaffolding proteins are clustered.  (4) As synapses mature, there is additional recruitment of presynaptic and postsynaptic proteins that leads to stabilization of the contact (Gerrow and El-Husseini, 2006).    5 The second step in synaptogenesis is thought to involve recruitment of pre- and postsynaptic components at the site of initial contact. This includes recruitment of presynaptic release machinery and postsynaptic neurotransmitter receptors, scaffolding proteins and associated signaling molecules. The transport of synaptic proteins has been shown to occur through several modalities, including, the delivery of transport packets (Ahmari et al., 2000; Bamji et al., 2003; Bresler et al., 2004; Gerrow, 2005; Prange and Murphy, 2001; Washbourne et al., 2002; Washbourne et al., 2004), lateral diffusion along the membrane from extra-synaptic sites (Dahan et al., 2003; Groc et al., 2003; Tardin et al., 2003), and gradual accumulation from a diffuse cytoplasmic pool (Bresler et al., 2001; Bresler et al., 2004). After recruitment of pre- and postsynaptic proteins, these newly formed synapses may be stabilized or lost, in an effort to fine tune neuronal circuitry.  Synapse formation requires the coordinated action of several presynaptic and postsynaptic proteins, and two main groups of proteins have emerged as key regulators, namely adhesion and scaffolding molecules. Cell adhesion molecules are membrane- anchored proteins involved in linking two cells together at points of contact. Each of the steps of synaptogenesis outlined herein requires cohesion and reciprocation between two cells, and this is a fundamental property of cell adhesion molecules.  Cell adhesion molecules have long been surmised to play a role in keeping synapses together, and recently have been shown to be important in other aspects such as triggering synapse formation (Biederer et al., 2002; Fu et al., 2003; Scheiffele et al., 2000), target recognition (Shen et al., 2004; Yamagata et al., 2002), and maturation (Chavis and Westbrook, 2001; Togashi et al., 2002). Scaffold molecules, are multi-modular proteins involved in the assembly of protein complexes. In presynaptic terminals, scaffolding proteins have been shown to be important in orchestrating the assembly of active zones and the recruitment of synaptic vesicles (Ahmari et al., 2000; Bamji et al., 2003; Bresler et al., 2004; El-Husseini et al., 2000a; El-Husseini et al., 2000b; Sala et al., 2001; Srivastava and Ziff, 1999; Zhai et al., 2001; Ziv and Garner, 2004). Scaffold proteins on the postsynaptic side have been shown to be important for the recruitment of receptors, and in the maturation of synaptic contacts (Ehrlich and Malinow, 2004; El-Husseini et  6 al., 2000a; El-Husseini et al., 2000b; Kneussel et al., 2001; Sala et al., 2001; Srivastava and Ziff, 1999; Stein et al., 2003; Takahashi et al., 2003).  1.2 DEVELOPMENT OF THE NEUROMUSCULAR JUNCTION  Much of our knowledge about synapse formation and maturation is derived from studies on the neuromuscular junction (NMJ), contacts formed between presynaptic motoneurons which contact postsynaptic muscle cells (Goda and Davis, 2003; Sanes and Lichtman, 1999). Due to the large size and accessibility of this synapse, much of its physiological properties and the molecules responsible are well characterized, as well as knock-outs generated for many of the proteins involved (Figure 1.3).      Figure 1.3: Molecular constituents at the NMJ. The nerve terminal occupies a shallow gutter in the muscle fiber and is capped by processes of Schwann cells. Active zones in the nerve terminal directly appose junctional folds in the postsynaptic membrane. Some of the proteins concentrated at the synapse are shown, with their subcellular localizations indicated by arrows. Those for which knockout mice have been generated are indicated in boldface. Reproduced with permission from (Sanes and Lichtman, 1999).          7 1.2.1 Pre-patterning  The earliest event in synapse formation at the NMJ is a rudimentary postsynaptic organization, termed pre-patterning, which occurs before arrival of the nerve axonal growth cone. In this pre-patterning, acetylcholine receptors (AChRs) concentrate on the muscle fiber, independent of contact with motoneuron terminals (Arber, 2002; Lin, 2001; Yang, 2001). The mechanisms underlying pre-patterning are unclear, however data obtained from knockout mice indicate that the function of muscle-specific kinase (MuSK) is required, and that this process is independent of the secreted factor agrin (Lin, 2001; Yang, 2001). The role of pre-patterning in subsequent synapse formation also remains unclear.  In vitro, synapse formation occurs when pre-patterning has not been observed, and incoming neurites do not preferentially contact pre-clustered AChR¶s (Anderson, 1977; Frank, 1979). Thus pre-patterning of AChR¶s is not required for subsequent synapse formation in vitro. However, in vivo AChR clusters are preferentially formed at the endplate band where innervation eventually occurs (Zhou, 1997), suggesting that the location of synapses may not be entirely determined by the site at which the motoneuron contacts the muscle.  Based on these findings, one may envision two scenarios: motor axons ignore these preformed clusters in vivo just as they do in vitro, and use agrin to organize new clusters and the use a µsecond signal¶ to disperse non-synaptic clusters. Alternatively, axons might recognize these clusters or encounter them by chance, and then use agrin to enlarge and/or stabilize them. The latter appears to be the case based on work in zebrafish, where motor axon growth cones and filopodia are selectively extended toward and contact pre- patterned AChR clusters, followed by the rapid clustering of presynaptic vesicles and insertion of additional AChRs (Panzer et al., 2006). Recent evidence has suggested that this may also be true in mammals, and that the tyrosine kinase ErB2 may help establish the distribution of pre-patterned AChRs (Vock et al., 2008).    8 1.2.2 Postsynaptic maturation and reciprocation  Upon arrival of the motoneuron terminal, secreted molecules such as neuregulin and agrin, are released from axonal terminals and induce further clustering of AChR and maturation of the postsynaptic specializations in the target muscle cells by acting through ErbB tyrosine kinase and MuSK (DeChiara, 1996; Glass et al., 1996; McMahan, 1990). The phenotype of the agrin and MuSK knockout mice provide evidence that this postsynaptic differentiation is necessary for subsequent induction of presynaptic development, as these motoneuron terminals fail to differentiate and remain highly dynamic (DeChiara, 1996; Gautam et al., 1996). The muscle-derived signal necessary for this reciprocal differentiation has not been defined, however, signaling via laminins have been implicated since knockout mice of several isoforms of laminin results in perturbed presynaptic development (Noakes et al., 1995; Patton et al., 2001; Sunderland et al., 2000). The downstream signaling mechanism of this reciprocal signal remains unclear.  1.2.3 Synaptic refinement and re-modeling  During the early stages of development, muscle fibers are innervated by several motoneurons (polyneural innervation). A few days later, around the time of birth, synapse elimination results in the separation of motor units, and establishment of the adult pattern of mononeural innervation. This process has been shown to be regulated by the limited availability of trophic support (Bennett and Robinson, 1989; Thompson, 1985), as well as motoneuron activity (Balice-Gordon and Lichtman, 1994; Ribchester and Taxt, 1983; Thompson, 1985). Together, the known interactions between the presynaptic motoneuron and postsynaptic muscle cell represent a core, albeit incomplete, understanding of the main steps in NMJ synaptogenesis. The emerging view is that the reciprocal signaling typified at the NMJ may be recapitulated for the formation of CNS synapses, but most likely with different molecular players.   9 1.3 DEVELOPMENT OF SYNAPSES IN THE CNS  In contrast to the NMJ, mechanisms of synaptogenesis in the central nervous system (CNS) are less understood (Craig and Boudin, 2001; Rao et al., 1998). This process is complicated by the enormous heterogeneity of the neuronal types and the differences in timing of their development. There are many types of synapses in the brain, each identified by the neurotransmitter they release. The effect of the neurotransmitter on the postsynaptic neuron is context specific, depending on several factors including the developmental stage of the neuron, the type and number of receptors present on the postsynaptic membrane, and neuronal activity (Wilson et al., 2004).  Synaptogenesis is an amazingly precise process considering that a single neuron may receive tens of thousands of heterogeneous synaptic inputs, yet virtually no mismatches between pre- and postsynaptic elements occur (Lardi-Studler and Fritschy, 2007). The main inhibitory neurotransmission in the vertebrate CNS is γ-aminobutyric acid (GABA), and is responsible for modulation of every aspect of brain function (the formation of these synapses will be discussed later). Excitatory synaptic transmission in the mammalian brain is primarily mediated by the neurotransmitter glutamate, and most of our knowledge of synaptogenesis in the CNS is based on research done on glutamatergic synapses. Due to lack of evidence to the contrary, events of glutamatergic synapse formation are presumed to be similar for other types of synapses found in the CNS, and is thought to involve the characteristic steps of contact initiation, recruitment of presynaptic and postsynaptic proteins, and stabilization. Both scaffolding proteins and adhesion complexes have been implicated in each of these steps.  1.3.1 Contact initiation  An important aspect of synaptogenesis is the ability of axons to grow to their target fields and synapse with the correct postsynaptic cell. These axons often travel long and arduous pathways before reaching their final target, and although they come in contact with a  10 multitude of potential postsynaptic partners along the way, they do not establish synapses on inappropriate cells. For instance, motoneuron axons from the ventral horn of the spinal cord delay synapse formation for days, even weeks, within their target regions (Sanes and Lichtman, 1999). Thus, precise axon guidance by cues alone is not sufficient for target- specific synapse formation and suggests that other factors must contribute (Vogt et al., 2005). Several target-derived factors have been identified, and many adhesion molecules have been implicated in this process (Gerrow and El-Husseini, 2006). Contact initiation has traditionally been described in terms of an axonal growth cone that sends out filopodia searching for a prospective partner. This view is likely a carry over from studies of the NMJ where the target muscle is relatively stationary. Dendrites also extend growth cones, and are lined with filopodia along their entire length (Figure 1.4). The dynamic protrusive behavior of these dendritic filopodia is consistent with an active ability to initiate synapse formation (Fiala et al., 1998; Ziv and Smith, 1996).   Figure 1.4: Initiation of synaptogenic contacts by axons and dendrites. Two axons (blue) and one dendrite (green) elongate, extend processes, and initiate contact sites that might evolve potentially into synapses. Principal events: 1, 2, interactions between axonal and dendritic growth cones; 3, an axonal growth cone intersecting with a dendritic shaft; 4, 5, dendritic filopodia contacting axonal shafts; and 6, a dendritic growth cone contacting an axonal shaft. Reproduced with permission form (Ziv and Garner, 2001).   Given the large number of neuronal contact types and targets, adhesion systems mediating initial contact are predicted to be complex and polymorphic in order to offer  11 sufficient combinatorial possibilities. This µlock and key¶ mechanism was which implies the existence of specific adhesion molecules that pair axons with their targets for specification of synaptic connections (Sperry, 1963). Cadherins are family of proteins shown to be important in contact initiation. In development, cadherins are distributed diffusely along the length of dendritic motile filopodia and upon contact with an axon, the cadherin complex accumulates at points of contact (Jontes et al., 2004; Togashi et al., 2002). Studies using a dominant negative approach, demonstrated that blockade of cadherin function results in a conversion of mature spines to a more immature structure via appearance of filopodia-like spines (Togashi et al., 2002). Interfering with cadherin function suggests that these molecules are more likely involved in regulation of target specificity, rather than synapse induction since initial synaptic assembly is delayed, but not blocked in neurons that are transfected with dominant-negative cadherin (Bozdagi et al., 2004; Inoue and Sanes, 1997). Thus, evidence indicates that cadherins are important for target recognition, but they are not essential for synapse differentiation which may require other molecular systems.  1.3.2 Recruitment of presynaptic and postsynaptic proteins  The second step in synaptogenesis involves recruitment of presynaptic and postsynaptic components at the site of initial contact. This includes recruitment of the presynaptic release machinery and postsynaptic neurotransmitter receptors and associated signaling molecules. Recruitment of different proteins to presynaptic and postsynaptic sites suggests involvement of heterotypic trans-synaptic signals. These trans-synaptic signals would require specificity, the ability to recruit the appropriate neurotransmitter on the presynaptic side and their cognate receptors on the postsynaptic side, in order to avoid mismatching. A sequential order for the recruitment of synaptic proteins and whether they are recruited as preassembled complexes or as individual molecules has been an area of intense study (reviewed in (McAllister, 2007; Waites et al., 2005a; Zhen and Jin, 2004), however, consensus on the hierarchal recruitment of individual proteins remains elusive, and multiple methods of assembly have been demonstrated (Figure 1.5).  12    Figure 1.5:  Multiple mechanisms for the recruitment and stabilization of synaptic proteins to new sites of contact. (a) Glutamatergic synapses between axon and dendrite shafts of hippocampal neurons can form in about an hour after the initial accumulation of presynaptic vesicles. Presynaptic proteins, including synaptic vesicle precursors (STVs) and piccolo-transport vesicles (PTVs), are mobile in axons before synapses are formed (first panel). These precursors are the first proteins recruited to nascent synapses (second panel). After 30 min, PSD-95 accumulates at these sites (third panel) followed by glutamate receptors (fourth panel)  13 (Bresler et al., 2004; Friedman et al., 2000). (b) In young cortical neurons, glutamatergic synapses can form even faster, on a timescale of several minutes. In these cells, STVs and NMDARs are both found in transport packets that are highly mobile in the axons and dendrites, respectively, before synapse formation (first panel). Both STVs and NMDAR transport packets cycle with the membrane during their transport (Sabo and McAllister, 2003; Washbourne et al., 2002). Contact between an axonal growth cone filopodium and a dendrite (right), or between axon and dendrite shafts (left), leads to the rapid and simultaneous recruitment of STVs and NMDARs at nascent synapses within 7 min of contact (second panel). PSD-95 is recruited to these sites with a variable time course, and AMPARs are recruited an hour following initial recruitment of NMDARs (third panel). (c) Glutamatergic synapses can also form at prespecified sites along dendritic shafts of hippocampal neurons, defined by stable preformed scaffold complexes associated with neuroligin. Complexes of scaffolding proteins (including PSD-95, Shank, and GKAP) are mobile within dendrites before synapses are formed (first panel). When these complexes associate with neuroligin, they often become stabilized in the dendritic membrane (second panel). These complexes can then recruit STVs to form synapses within 2 h of their stabilization (third panel)(Gerrow, 2005). (d) There are also predefined sites along the axon shaft of cortical neurons where en passant synapses selectively form. These predefined sites are stable locations along the axon where STVs cycle with the plasma membrane (first panel) and presumably release diffusible molecules before synapses are formed (second panel). Filopodia from dendritic growth cones (right) and presumably also dendritic shafts (left) are selectively attracted to, and stabilized at, these sites (third panel). Following stabilization of this contact at this predefined site, the presynaptic terminal is formed and additional pre- and postsynaptic proteins are recruited to form a nascent synapse (fourth panel) (Sabo et al., 2006). Reproduced with permission from (McAllister, 2007).   1.3.2.1 Presynaptic recruitment The major events in presynaptic terminal differentiation are the formation of the active zone and the clustering of synaptic vesicles. A number of proteins that are present in the presynaptic active zone have been identified. Time-lapse imaging studies have captured dynamic and transient events in the transport of presynaptic components, and therefore provided insight into the early stages of synaptogenesis. Local recruitment of individual molecules probably contributes to presynaptic assembly, but studies from several laboratories have suggested that vesicular delivery plays a prominent role. The fusion of dense core vesicles carrying structural components of the presynaptic active zone have been observed shortly after initial contact (Ziv and Garner, 2004). These vesicles carry scaffolding proteins (Hannah et al., 1999; Lee et al., 2003; Matteoli et al., 1992; Ohtsuka et al., 2002; Shapira et al., 2003) as well as components of the synaptic vesicle exocytotic machinery (Shapira et al., 2003; Zhai et al., 2001). Another population of precursor presynaptic vesicles containing proteins important for active neurotransmitter release, such as VAMP and synaptophysin, has also been observed to fuse shortly after initial contact (Ahmari et al., 2000).  14 The order and timing of recruitment of these two populations of presynaptic transport vesicles has not been directly compared, and the molecular cascades responsible for their recruitment are unknown. Cell adhesion complexes have often been implicated, however, the order and delivery mode for most adhesion molecules during synapse formation has not determined. For instance, neurexin and SynCAM are potent inducers of presynaptic active zone assembly onto heterologous cells, and it can be presumed that their delivery is an early event resulting in the transformation of a nascent contact to a functional presynaptic zone. SynCAM and neuroligins are the only CAMs so far to demonstrate the ability to promote presynaptic differentiation at contacts between axons and heterologous cells, and that drive the recruitment of synaptic proteins at sites of contact (Biederer et al., 2002; Dean et al., 2003; Scheiffele et al., 2000). This suggests that at least some cell adhesion molecules are sufficient to drive recruitment of synaptic proteins to contact sites. However, it is unclear whether these molecules are already at the plasma membrane and cluster through lateral movement, or whether they are directly delivered via vesicles.  1.3.2.2 Postsynaptic recruitment One of the earliest events in postsynaptic differentiation is the recruitment of scaffolding proteins of the PSD-95 family. These molecules are present at synapses in postnatal day 2 hippocampus (Sans et al., 2000), and detectable within 20 min of axodendritic contact in culture (Bresler et al., 2001; Friedman et al., 2000; Okabe et al., 2001). Gradual accumulation of PSD-95 could occur by local trapping of diffuse plasma membrane pools, or by sequential local fusion of numerous vesicles, each carrying only small numbers of PSD-95. Some investigators have reported transport of recombinant PSD-95 clusters during synaptogenesis (Gerrow, 2005; Prange and Murphy, 2001), whereas other studies have observed more gradual accumulation of PSD-95 at nascent synapses (Bresler et al., 2001; Bresler et al., 2004; Marrs et al., 2001). Closely following the synaptic recruitment of PSD-95 is recruitment of NMDA-type and AMPA-type glutamate receptors, which are independently regulated. As described above for PSD-95, transport of NMDA receptor clusters during synaptogenesis has been reported (Washbourne et al., 2002), and other studies have observed more gradual  15 accumulation at nascent synapses (Bresler et al., 2004). For AMPA receptors, evidence exists to support both local insertion of receptor-containing vesicles near the synapse, and insertion over the bulk of the dendritic plasma membrane followed by diffusion and trapping at the synapse (Bats el al., 2007; Borgdorff and Choquet, 2002; Passafaro et al., 2001). Much effort has focused on how synaptic delivery of AMPA and NMDA receptors is controlled by interacting proteins (Figure 1.6, reviewed in (Bredt and Nicoll, 2003; Malinow and Malenka, 2002; Wenthold et al., 2003)). Classically, synaptic clustering of glutamate receptors has been shown to be regulated through direct and indirect interactions with scaffolding proteins, and in particular with scaffolding proteins containing PDZ domains ( Kornau et al., 1995; Chen et al., 2000; El- Husseini et al., 2000a; Kim and Sheng, 2004). The clustering of glutamate receptors at synaptic sites has also been shown to be influenced by adhesion complexes (Dalva et al., 2000; Passafaro et al., 2003; Kayser et al., 2006; Saglietti et al., 2007). Using the reconstituted synapse, Huganir and colleagues identified the minimum number of proteins sufficient for AMPA receptor recruitment to excitatory contact sites: an adhesion complex containing neuroligin, and the secreted factor NP-1 and its receptor (Sia et al., 2007). Figure 1.6: Different modes that regulate clustering of AMPA receptors at the synapse.  (1) Postsynaptic scaffolding proteins such as PICK, GRIP, and SAP-97 directly bind to the C-terminal PDZ-binding motif of AMPA receptors (AMPAR). (2) Coupling of scaffolding proteins such as PSD-95 to TARPs controls AMPA receptor clustering and retention at the synapse. This also provides a link between AMPA receptors and adhesion complexes, such as neuroligin/neurexin. (3) Adhesion molecules such as N-cadherin (N-CAD) bind directly to the extracellular domain of AMPA receptors to influence their clustering. (4) Secreted factors from the presynaptic terminal such as the pentraxin Narp interact with the AMPA receptor extracellular domain to promote clustering in an activity-dependent manner. (5) The secreted pentraxin NP-1, associates with NPR to induce clustering of AMPA receptors. Reproduced with permission from (Gerrow and El-Husseini, 2007).   16 1.3.3 Maturation  After recruitment of presynaptic and postsynaptic proteins, these newly formed synapses may be lost or further stabilized, in an effort to fine tune neuronal circuitry. A feature of synaptic development is a prolonged maturation phase, where the number of synaptic vesicles in the presynaptic compartment increases, as well as at the size and protein content of the post synaptic density (Pierce and Mendell, 1993; Vaughn, 1989). The most dramatic example of synapse maturation is the transformation of an immature glutamatergic postsynaptic compartment into a mature spine, an actin-based protrusion with a bulbous head extending from the dendrites (Figure 1.7). By having an enlarged head attached to the dendritic shaft by a small neck, spines provide biochemical as well as spatial compartmentalization. This morphological change of the postsynaptic compartment is regulated by numerous mechanisms that signal through proteins that regulate the actin cytoskeleton such as the Rho family of GTPases (Ethell and Pasquale, 2005).   Figure 1.7: Dendritic spine morphology and signaling. (A) Dendritic tree of a GFP-labeled hippocampal neuron with many dendritic spines. (B) Higher magnification view shows that most dendritic spines have a narrow neck and an enlarged head. (C) A hippocampal neuron labeled with GFP-actin showing accumulation at spine heads. (D) GFP-labeled dendrites (green) with spine heads labeled with actin (red). (E) Signaling cascades that regulate dendritic spine shape and motility involve cell surface receptors and ion channels, which activate signaling cascades controlling the activity of Rho GTPases.  Reproduced with permission from (Ethell and Pasquale, 2005).   17 In addition to their adhesive properties, many CAMs interact with proteins that affect the actin cytoskeleton, and thus are poised to mediate this maturational change (Carlisle and Kennedy, 2005; Hering and Sheng, 2001; Ethell and Pasquale, 2005). For instance, nectins are members of the immunoglobulin super-family which functions in concert with cadherins in various types of adherens junctions including the puncta adherentia of synapses, can affect the actin cytoskeleton via activation of Rac1 and Cdc42 (Kawakatsu et al., 2002; Mizoguchi et al., 2002). The scaffolding protein Shank is also a potent inducer of spines, and requires interaction another scaffolding protein Homer (Tu, Xiao et al. 1999; Sala, Piech et al. 2001). Although both adhesion complexes and scaffolding proteins have been implicated in spine formation, many of the molecular cascades responsible remain unknown.  1.3.4 Elimination  Although not required for initial steps of synapse formation (Rao and Craig, 1997), the process of synapse elimination or strengthening is thought to, in part, depend on external stimuli and synaptic activity. For instance, in the rodent barrel cortex which receives somatosensory input from the whiskers, sensory deprivation via trimming of the whiskers (loss of external stimuli), results in decreased stable synaptic connections (Trachtenberg et al., 2002). In contrast, stimulation of the whisker has been shown to increase the number of synapses (Knott et al., 2002). Therefore, cell adhesion systems important for these processes are likely to be modulated by synaptic activity leading to changes in protein levels or in their adhesive properties.  Most work has focused on the processes of synapse formation and maturation, however, synapse elimination is equally an important developmental process (Goda and Davis, 2003). Initial numbers of synapses in development are greater than the number retained into adulthood, highlighting that synapse elimination is a crucial step in normal brain development (Hashimoto and Kano, 2003; Lichtman and Colman, 2000). The pruning of synapses is a common theme during early activity-dependent refinement of neuronal  18 circuitry. There are two proposed mechanisms of synapse removal: Input elimination and synapse disassembly. During input elimination, a presynaptic cell loses all synaptic contacts with a postsynaptic target, functionally and anatomically uncoupling the cells although contact with other targets persist, and has been described in several neuronal circuits (Keller-Peck et al., 2001; Hashimoto and Kano, 2003; Jackson and Parks, 1982; Sanes and Lichtman, 1999; Sretavan and Shatz, 1986). In contrast, synapse disassembly refers to the deconstruction of individual synaptic contacts, and thus represents a mechanism for modulating the strength of connectivity between two cells. This has been observed at the NMJ and CNS synapses in invertebrates (Goda and Davis, 2003). The molecular mechanisms that drive synapse disassembly and input elimination remain unclear. Examination of fixed preparations suggest that presynaptic elimination of synapsin and vesicle associated proteins precede the removal of postsynaptic receptors (Eaton et al., 2002; Hopf et al., 2002). The speed of synapse disassembly can be substantially faster than the rate of protein turnover, indicating destabilizing mechanisms rather than simple removal of trophic support (Huh and Wenthold, 1999).  One would predict that loss of adhesion would play an integral part of synapse destabilization. This has been demonstrated for long term depression (LTD) in invertbrates (Zhu et al., 1995), and at the NMJ where matrix metalloproteases are able to remove agrin and ACh receptors from the postsynaptic membrane (Tyc and Vrbova, 1995; VanSaun et al., 2003). Whether proteases work similarly at CNS synapses to cleave adhesion complexes and initiate synapse elimination remains to be determined. However, in HEK293 or COS cells transiently or stably over-expressing the adhesion molecule protocadherin- (Pcdh-̛̓, it was found that the -Pcdh ectodomain was cleaved at the cell surface, and that this cleavage could be blocked by inhibitors of matrix metalloproteinases (Haas et al., 2005; Hambsch et al., 2005). These results suggest a very interesting paradigm where the cleavage of adhesion molecules may play an important role in disengaging synaptic contacts.    19 1.4 PSD SCAFFOLDING PROTEINS  Glutamatergic synapses in the CNS are characterized by an electron dense web underneath the postsynaptic membrane called the postsynaptic density (PSD). PSDs contain a dense network of hundreds of proteins, including cell adhesion, cytoskeletal proteins, adaptor and scaffolding proteins, membrane-bound receptors and channels, g- proteins and signaling molecules including kinases and phosphotases (Kennedy, 1997; Sheng, 1996; Sheng and Sala, 2001; Ziff, 1997). Identification of the myriad of PSD proteins has been significantly aided by proteomics analysis of biochemically purified synaptic preparations (Figure 1.8)(Peng et al., 2004; Walikonis et al., 2000). And recently, topographic reconstructions from freeze-substituted hippocampal cultures provide the first views of the three-dimensional molecular organization of the PSDs in vivo (Chen et al., 2008b).   Figure 1.8: Proteins from postsynaptic density (PSD) fractions. The synaptosome fraction and PSD fractions that were obtained after washing with the indicated combinations of detergents were fractionated by SDS-polyacrylamide gel electrophoresis on a 7% acrylamide gel. Proteins were stained with Coomassie blue. Identified proteins are indicated on the right. Positions of molecular weight markers are indicated on the left. Abbreviations: Syn, synaptosomes; 1 Triton, the pellet obtained after one extraction of synaptosomes with Triton X-100. Reproduced with permission from (Kennedy, 1997).     The molecular architecture of the PSD is largely built around PDZ containing scaffold proteins, and in particular the PSD-95 (also known as SAP90) family of MAGUK proteins. The MAGUK (membrane-associated guanylate kinases) family of synaptic proteins share a common domain organization, with one or three N-terminal PDZ (PSD- 95, Dlg and ZO-1) domains, an SH3 domain, and a C-terminal region homologous to  20 guanylate kinases. PDZ domains mediate protein-protein interactions and typically bind to short amino acid motifs at the C-termini of interacting proteins that include certain ion channels and receptors (Kim and Sheng, 2004). A theme that emerges from recent findings is that stoichiometry between interacting proteins (such as receptors) and scaffolding proteins may regulate their targeting to and/or retention at contact sites. Whether scaffolding molecules cooperate or compete with one another to regulate the number of interacting molecules at the synapse remains to be determined. Competitive interactions may be determined by differential affinities of postsynaptic scaffolding proteins for certain molecules, whereas scaffolding proteins present at presynaptic terminals may then indirectly cooperate with postsynaptic scaffolding proteins to stabilize synapses (Montgomery et al., 2004). The role of scaffold proteins in modulating adhesion is an attractive hypothesis since they may represent the vehicle for crosstalk between different sets of adhesion complexes, thus coordinating the multiple proteins and events of synaptogenesis.  1.4.1 Protein interactions of PSD-95  The role of PSD-95 in the organization of the PSD was first uncovered through the discovery that NMDA receptors bind to PSD-95. NMDA receptors are composed of NR1 and NR2 subunits, the latter having long cytoplasmic tail with a conserved C-terminal sequence (íESDV or íESEV) that binds to the first two PDZ domains of the PSD-95 family (Kornau et al 1995, Niethammer et al 1996; Muller et al., 1996). Since then, many protein interactions have been defined for PSD-95, making it known as a ³master scaffold´ at glutamatergic synapses (Figure1.9)(Kim and Sheng, 2004).  21   Figure 1.9: Schematic diagram of PDZ proteins at a mammalian excitatory synapse. The most studied PDZ-containing proteins of a glutamatergic synapse are shown, only a subset of known protein interactions is illustrated, as indicated by the overlap of proteins. Green and blue ellipses in PSD-95 represent SH3 and GK domains, respectively. Crooked lines indicate palmitoylation of PSD-95 and GRIP. Grey arrows indicate binding and/or regulatory actions of proteins on the actin cytoskeleton. AKAP79, A- kinase anchor protein 79; AMPAR, AMPA (Į-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptor; ȕPIX, PAAK-interactive exchange factor; CaMKIIĮ, Į-subunit of Ca2+/calmodulin-dependent protein kinase II; GK, guanylate kinase-like domain; EphR, ephrin receptor; ErbB2, EGF-related peptide receptor; GKAP, guanylate kinase-associated protein; GRIP, glutamate-receptor-interacting protein; IP3R, IP3 receptor; IRSp53, insulin-receptor substrate p53; K ch, potassium channel; LIN7, lin7 homologue; LIN10, lin10 homologue; mGluR, metabotropic glutamate receptor; NMDAR, NMDA (N-methyl-D- aspartate) receptor; nNOS, neuronal nitric oxide synthase; PICK1, protein interacting with C kinase 1; PSD-95, postsynaptic density protein 95; SER, smooth endoplasmic reticulum; SH3, Src homology 3 domain; Shank, SH3 and ankyrin repeat-containing protein; SPAR, spine-associated RapGAP; SynGAP, synaptic Ras GTPase-activating protein. Reproduced with permission from (Kim and Sheng, 2004).   Receptors and channels. Additional receptors that can bind to PSD-95 via C terminus± PDZ interactions include voltage-gated K+ channels (Kim et al 1995), inward rectifying K+ channels (Cohen et al. 1996, Hibino et al. 2000, Nehring et al. 2000), plasma membrane calcium pumps (Kim et al. 1998), and kainate-type glutamate receptors (Garcia et al 1998). In initial screens, AMPA-type glutamate receptors, important for,  22 were generally not associated with PSD-95; and interacted instead with other scaffold proteins, namely SAP97, GRIP and ABP (Dong et al 1997; Leonard et al 1998; Srivastava et al 1998). PSD-95 influences AMPA receptor recruitment to synapses through an indirect mechanism (El-Husseini et al., 2000a) that was not understood until the discovery of stargazin, a transmembrane protein that directly interacts with AMPA receptors.  Stargazin, and other members of the TARP (transmembrane AMPA receptor- binding protein) family, are expressed throughout the brain, and provide the link through which PSD-95 facilitates synaptic clustering of AMPA receptors (Chen et al., 2000).  Adhesion molecules. Neuroligin, a postsynaptic transmembrane ligand for presynaptic neurexin, binds to PDZ3 of PSD-95 (Irie et al., 1997, Song et al., 1999). The C-terminal of cadherins interacts with A-kinase-anchoring protein (AKAP) 79/150. It has been shown that AKAP organizes a scaffold of cAMP-dependent protein kinase (PKA), protein kinase C (PKC), and protein phosphatase 2B/calcineurin that regulates phosphorylation pathways underlying neuronal LTP and long-term depression (LTD) synaptic plasticity. (Gorski et al., 2005).  Synaptic adhesion-like molecules (SALMs) have a single transmembrane (TM) domain and contain extracellular leucine-rich repeats, an Ig domain, a fibronectin type III domain, and an intracellular PDZ binding domain. SALM1 interacts with PSD-95, as well as other MAGUKs including SAP102 and SAP97 (Wang et al., 2006).  Signaling proteins. ErbB4, a receptor tyrosine kinase for neuregulin, binds to PDZ1/2 of PSD-95 (Garcia et al. 2000, Huang et al. 2000). PSD-95 binds to nNOS via a PDZ±β- finger interaction (Brenman et al. 1996a). Regulators or effectors of small GTPases Ras, Rho, and Rap have been found to bind to PSD-95 and to be associated with the PSD (Sheng&Pak 2000). The most prominent of these is SynGAP, a GTPase activating protein for Ras that interacts with all three PDZ domains of PSD-95 (Chen et al. 1998, Kim et al. 1998). PSD-95 interacts with Fyn, via binding of the Fyn SH2 domain to PDZ3 of PSD-95, thereby enhancing the phosphorylation of NMDA receptors by Fyn (Tezuka et al. 1999).   23 Cytoskeletal proteins. Band 4.1, an actin/spectrin-binding protein of the ezrin-radixin- moesin family, binds in vitro to SAP97 and possibly to other PSD-95 family members (Lue et al. 1994, 1996; Marfatia et al. 1996, Wu et al. 1998). Such an interaction has the potential to link the PSD-95 complex to F-actin, which is the predominant cytoskeleton of dendritic spines. PSD-95 also interacts with microtubule-associated proteins: the PDZ3 of PSD-95 binds to CRIPT (Niethammer et al. 1998, Passafaro et al. 1999) and MAP1A via the GK domain (Brenman et al. 1998). The finding that PSD-95 interacts with microtubule binding proteins, thus linking the PSD to the tubulin-based cytoskeleton, is surprising since microtubules are sparse or absent from dendritic spines. However, microtubule anchoring may be relevant for the minority of excitatory synapses that exist on dendritic shafts where microtubules are more abundant, or these microtubule interactions may someway be involved in trafficking of the PSD-95.   Other scaffolding proteins. The GK domain binds to GKAP (also known as SAPAP or DAP) (Kim et al. 1997, Naisbitt et al. 1997, Satoh et al. 1997, Takeuchi et al. 1997) and BEGAIN (Deguchi et al. 1998). GKAP in turn binds to the Shank family of PDZ- containing scaffold proteins, thereby greatly increasing the extent and complexity of the protein network of PSD-95 (Naisbitt et al. 1999, Sheng & Kim 2000, Tu et al. 1999). Because Shank interacts with Homer, a metabotropic glutamate receptor-binding protein, the PSD-95±GKAP±Shank chain of interactions could link NMDA and AMPA receptors to metabotropic glutamate receptors in the postsynaptic membrane (Sheng & Kim 2000).  1.4.2 Other PDZ proteins  In addition to PSD-95, several other PDZ proteins have prominent roles in the establishment and maturation of synapses (Table 1.1).  GRIP (glutamate receptor interacting protein) is a 7-PDZ domain-containing protein () belonging to a family of highly homologous proteins that includes GRIP1, GRIP2, and the AMPA receptor-binding protein ABP (Dong et al., 1997; Dong et al., 1999;  24 Srivastava et al., 1998; Srivastava and Ziff, 1999; Wyszynski et al., 2002). GRIP can dimerize, and is present in both axons and dendrites (Wyszynski et al., 2002). GRIP interacts with many synaptic proteins: including AMPA receptors (Dong et al., 1997; Srivastava et al., 1998; Wyszynski et al., 2002); Eph receptors and their ephrin ligands (Hoogenraad et al., 2005); a RAS guanine nucleotide exchange factor (Ye et al., 2000); liprin-α (Wyszynski et al., 2002); and metabotropic and kainate-type glutamate receptors (Hirbec et al., 2002). The importance of GRIP is supported by evidence from knockout mice in which the Grip1 gene is disrupted, and show embryonic lethality (Bladt et al., 2002). Certain splice variants of GRIP can be palmitoylated, like PSD-95 and PSD-93, where palmitoylation of GRIP results in association with the plasma membrane at synapses and non-palmitoylated GRIP mostly associates with intracellular membranes (DeSouza et al., 2002). These differentially modified subpopulations of GRIP might stabilize synaptic AMPARs and be involved in the trafficking of intracellular pools respectively.  PICK is present at synaptic and non-synaptic sites in neurons, and its PDZ domain shows promiscuous binding. In addition to PKCα and GluR2/3, it has many other binding partners (both presynaptic and postsynaptic), including the netrin receptors (Williams et al., 2003), various metabotropic glutamate receptor subtypes (Perroy et al., 2002), and ErbB2 receptor tyrosine kinase (Hirbec et al., 2002). In many of these cases, the interaction with PICK1 seems to regulate the subcellular localization and/or surface expression of its protein partners. Phosphorylation of the C-terminus of GluR2 alters its binding specificity for GRIP and PICK1, and contributes to synaptic plasticity by altering the trafficking of AMPARs (Kim et al., 2001; Lin and Huganir, 2007). PICK1-knockout mice are viable and show normal synaptic transmission in several brain areas. However, cerebellar LTD is abolished, and can be rescued by transient transfection of PICK1- deficient Purkinje cells with wild type PICK1, but not mutants of PICK1 with mutations in the PDZ domains (Steinberg et al., 2006).  S-SCAM , also known as MAGI-2, is similar to PSD-95 in a domain organization but inverse structure: with an NH2-terminal GK-like domain followed by two WW and five  25 PDZ domains. It binds GKAP through the GK-like domain and NMDA receptors and neuroligins through the PDZ domains (Hirao et al., 1998). ȕ-catenin, a primary partner for classic cadherins, binds S-SCAM, and is involved in synaptic localization of S-SCAM (Nishimura et al., 2002). S-SCAM can also participate in the trafficking of AMPA receptors through PDZ-dependent interactions with TARPs (Deng et al., 2006). Interestingly, S-SCAM is localized at inhibitory synapses in rat primary cultured hippocampal neurons, and associates with beta-dystroglycan at these synapses (Sumita et al., 2007). Mutant mice lacking the slice variant S-SCAM alpha, have abnormally elongated dendritic spines that cannot be rescued by other splice variants, and is a scaffold required to activate RhoA protein in response to NMDA receptor signaling in dendrites (Iida et al., 2007).  Table 1.1: Other synaptic PDZ proteins PDZ PROTEIN FUNCTION INTERACTING PROTEINS REFS Afadin Involved in synapse adhesion and development Nectin, F-actin Eph receptors (receptor tyrosine kinases) (Takai and Nakanishi, 2003) Densin-180 Member of leucine-rich repeat and PDZ family CaM kinase IIα α-Actinin,  δ-Catenin (Walikonis et al., 2001) Erbin Suppresses the Ras±MAPK signaling pathway ErbB2 (receptor tyrosine kinase for neuregulin) PSD-95, δ-Catenin (Kolch, 2003) GRIP/APB AMPA receptor regulation and localization AMPARs, EpbB RasGEF, liprin-alpha/SYD2 (Dong et al., 1997; Srivastava and Ziff, 1999) Neurabin Modulates transmission and spine morphology Protein phosphatase 1 F-actin (Feng et al., 2000) PICK Glutamate receptor trafficking and surface expression PKCα, AMPAR, mGluR. Netrin receptors, ErbB3 (Xia et al., 1999) S-SCAM Regulate s assembly and trafficking of synaptic proteins NMDAR, Neuroligin KIF1Bα, β-Catenin , nRapGEF (Hirao et al., 1998; Ide et al., 1999; Nishimura et al., 2002) Shank Promotes morphological and functional maturation GKAP, Homer, Cortactin, CIRL, IRSp53, ABP1 βPIX,  Sharpin (Sala et al., 2001; Tu et al., 1999) Syntenin Small scaffold  AMPA, kainite and metabotropic receptors  Syndecan, Neurexin, SynCAM, ephrin B, (Hirbec et al., 2002) Tamalin Trafficking of mGLURs Group I mGluR, Cytohesin  GKAP, S-SCAM (Kitano et al., 2003)  Only proteins that directly interact with the indicated PDZ proteins are described. Owing to space limitations, this list is not comprehensive and not all relevant references are cited. AMPA, Į-amino-3- hydroxy-5-methyl-4-isoxazole propionic acid; Cdc42, Rac, Rap and Ras, small monomeric G- proteins;GKAP, guanylate kinase-associated protein; KIF1BĮ, kinesin family member 1BĮ; NMDAR, N- methyl-D-aspartate receptor; PSD-95, postsynaptic density protein 95; Ras±MAPK, Ras mitogen activated protein kinase; S-SCAM, synaptic scaffolding molecule.  26 1.4.3 Role in synapse formation  PSD-95 has been implicated in synapse development since it clusters at synapse precursors before many other postsynaptic proteins (Rao et al., 1998), and because discs- large (DLG), a PSD-95 homolog in Drosophila, is necessary for proper development of larval neuromuscular junctions (Lahey et al., 1994). At CNS glutamatergic synapses, PSD-95 is able to mediate changes in postsynaptic components which are accompanied by enhanced presynaptic maturation, indicating that PSD-95 induces the assembly of necessary molecules at the PSD to co-ordinate the maturation of pre- and postsynaptic elements (El-Husseini et al., 2000a). Indeed, mutant mice lacking PSD-95, the frequency function of NMDA-dependent LTP and LTD is shifted to produce strikingly enhanced LTP at different frequencies of synaptic stimulation. In keeping with neural-network models that incorporate bidirectional learning rules, this frequency shift is accompanied by severely impaired spatial learning (Migaud et al., 1998). The requirement for PDZ domains in this process suggests the involvement of PDZ-interacting proteins in synapse maturation. PSD-95 may be able to mediate this cross-synaptic maturation by the recruitment of adhesion molecules. At the PSD, neuroligin associates with two MAGUKS through its C-terminal PDZ-binding site: the third PDZ domain of PSD-95, and the first PDZ-binding domain of S-SCAM (Iida et al., 2004; Irie et al., 1997). The interaction between neuroligin and these scaffolding proteins may co-ordinate recruitment of other synaptic proteins.     27 1.5 NEUROLIGINS AND NEUREXINS  Neuroligins (NL) are postsynaptic transmembrane adhesion proteins whose extracellular domain associates with presynaptic partners, neurexins (NRX). Since the neuroligin- neurexin trans-synaptic complex was one of the first protein complexes determined to be sufficient to drive the formation of synapses, it has been an area of intense study (reviewed in (Biederer, 2005; Dean and Dresbach, 2006; Levinson and El-Husseini, 2005a; Levinson and El-Husseini, 2005b; Lisé and El-Husseini, 2006; Missler et al., 1998).  1.5.1 Structure of neuroligins  Members of the NL family are type I transmembrane proteins, are comprised of several domains, including a cleaved signal peptide, a cholinesterase-like domain, a carbohydrate attachment region, a single transmembrane domain and a short C-terminal tail containing a type I PDZ-binding motif  (Figure 1.10)(Ichtchenko et al., 1995). NL proteins have been identified in humans, rodents, chicken, Drosophila melanogaster and Caenorhabditis elegans (Bolliger et al., 2001; Gilbert and Auld, 2005; Ichtchenko et al., 1995; Ichtchenko et al., 1996; Kwon et al., 2004; Paraoanu et al., 2006). Three genes encoding NL family members have been identified in rat and mouse, while five genes coding for NLs have been identified in the human genome (Bolliger et al., 2001; Ichtchenko et al., 1996). Analysis of currently identified members shows that these proteins share 52% sequence identity. Intracellular regions are less conserved (31%) than extracellular (55%) and transmembrane domains (91% identity) (Ichtchenko et al., 1996).   28   Figure 1.10: Domain structure of neuroligins. (A) Members of the NL family are type I transmembrane proteins, comprised of several domains, including a cleaved signal peptide, a cholinesterase-like domain, a carbohydrate attachment region, a single transmembrane domain, and a short C-terminal tail containing a type I PDZ binding motif. Analysis of the currently identified members shows that these proteins share 52% sequence identity. The intracellular regions  showing 31% identity is less conserved than the extracellular and transmembrane domains, which show 55% and 91% identity, respectively. Inclusion of short alternative exons (20 and 9 amino acids in length) at two sites (designated A and B, respectively) generates four potential NL1 variants. NL2 and NL3 undergo a similar splicing event at the first site, thereby generating two different isoforms.   The extracellular region of NLs are responsible for heterophilic adhesion, and contains a domain with sequence similarity to cholinesterases, members of the α/β-hydrolase fold superfamily of enzymes. Because of this similarity, NLs belong to a family of adhesion molecules called cholinesterase-like adhesion molecules (CLAMs), which include glutactin, neurotactin and gliotactin (Gilbert and Auld, 2005; Scholl and Scheiffele, 2003). But unlike cholinesterases, NLs and their related family members, lack one residue in the catalytic triad located within the cholinesterase-like domain (CLD), which renders them enzymatically inactive. Within their CLD-domains, the human NL-1, -2, -3, and -4 isoforms exhibit nearly 70% amino acid identity. Nevertheless, the isoforms exhibit differential interactions with neurexin isoforms and differential localization at neuronal synapses (Budreck and Scheiffele, 2007; Song et al., 1999; Varoqueaux et al., 2004). Some of these distinct properties are determined by small alternative splice insertions at two sites within the CLD-domain termed A and B. NL2 and NL3 undergo alternative splicing only at the A site (17-aa insertion for NL2), and NL1 can additionally be modified by inclusion or exclusion of a 9-aa insertion at site B, thus generating four potential NL1 variants: NL1(í), NL1A, NL1B, and NL1AB (Ichtchenko et al., 1995).  29 1.5.2 Structure of neurexins  Functions of NLs at the synapse have been linked to their interaction with NRXs, a family of highly polymorphic brain-specific proteins identified through a search for receptors of the black widow spider toxin, α-latrotoxin (Ushkaryov et al., 1994). Mammalian NRXs are the product of three genes, referred to as NRX I, II and III, from which a long mRNA encoding α-NRX and a short mRNA encoding β-NRX are generated (Ullrich et al., 1995; Ushkaryov et al., 1994). Five canonical alternative splice sites in α-NRX, and two for β-NRX have been identified (Ullrich et al., 1995). Thus, alternative splicing of NRXs can potentially give rise to more than a thousand different transcripts (Missler and Sudhof, 1998). This extensive splicing represents a powerful mechanism for producing a multitude of distinct adhesion proteins that can be expressed within a single cell or population of cells, offering the diversity and specificity required for cell-cell recognition.  NRXs are classical single pass transmembrane proteins, with their N-terminal facing extracellular and their C-terminals intracellular (Figure 1.11). The N-terminal, or binding domain, of α-NRX is made of three homologous modules made of two LNS (Laminin, Neurexin, Sex-hormone-binding protein) domains organized with a single EGF (Epidermal Growth Factor)-like domains, which share high conservation with the LNS domain of agrin, including in the position of the splice sites (Rudenko 1999). β-NRXs have only one of these modules. Intracellularly, NRXs are highly conserved, and bind to several proteins including: synaptotagmin (Hata 1993); CASK (Hata et al., 1996), a protein of the MAGUK family that connects NRXs  to the exocytotic machinery and to the actin cytoskeleton (Butz et al., 1998; Biederer et al., 2000; Grootjaans et al. 2000; Biederer and Sudhof, 2001). These interactions constitute a link between NRXs and synaptic vesicles and the vesicle fusion apparatus in the presynaptic terminal.   30  Figure 1.11: Neurexin domain structure. Each NRX gene uses an upstream promoter to generate the larger Į-NRX and a downstream promoter to generate the smaller ȕ-NRX.. In Į-NRX, the LNS (laminin, neurexin, sex-hormone-binding protein) domains are organized with EGF (epidermal growth-factor)-like domains into three homologous modules, I±III. The position of each of five sites of alternative splicing (SS1±SS5) is indicated.  Adapted and reproduced with permission from (Craig and Kang, 2007).   1.5.3 Crystal structure and binding  The crystal structures reported for NL and β-NRX have revealed key features that modulate their trans-synaptic interaction (Araç et al., 2007; Chen et al., 2008a; Comoletti et al., 2007; Fabrichny et al., 2007; Koehnke et al., 2008). Importantly, the structure of these molecules as a complex helped elucidate how association may influence synaptic maturation and strength. From these studies, it is apparent that NLs exist as constitutive dimers: two helices from each monomer form a four helix bundle interface, bond together by hydrophobic interactions with a few hydrogen bonds (Figure 1.12)(Araç et al., 2007; Fabrichny et al., 2007; Koehnke et al., 2008). These data agree with previous functional studies where mutations in this region can disrupt dimerization (Comoletti et al., 2003; Dean et al., 2003). Interestingly, the Cys-loop of NLs was shown to display flexibility relative to that of the AChEs, potentially causing a looser connection between the molecular core and the four-helix bundle involved in dimerization of NL molecules  31 (Fabrichny et al., 2007). Whether this flexibility is important for regulating adhesive properties is unknown.    Figure 1.12: Structure of neuroligin-1. (A) Ribbon diagram of a NL1 protomer. Two views are shown related by a 180ƒ rotation around the specified axis. Į-helices are colored orange and ȕ sheets are colored cyan. The unique disulfide bond is colored magenta. (B) Ribbon diagram of the NL1 dimerization interface. Residues involved in hydrophobic interactions are shown in sticks and H bonds are shown by dashed lines. The two NL1 protomers are colored orange and magenta, respectively. Reproduced with permission from (Araç et al., 2007).   The crystal structure of the NL/NRX complex demonstrated the arrangement of two NRX monomers bound to identical surfaces on opposite sides of the NL dimer (Figure 1.13) (Araç et al., 2007; Chen et al., 2008a; Fabrichny et al., 2007). The binding interface is very flat, is consistent with a low-affinity complex, and corresponds to only a small proportion of their molecular surfaces (2.5% of NL and 7% of Nrx) (Fabrichny et al., 2007). Earlier studies showed the importance of calcium ions in the binding of NL to Nxn (Ichtchenko et al., 1995; Ichtchenko et al., 1996; Nguyen and Sudhof, 1997), and modeling suggested that the ability of NL to bind Ca2+ might be based on two EF-hand like motifs (Tsigelny et al., 2000). This modeling data could not be recapitulated in the crystallization of NL alone (Koehnke et al., 2008). In contrast, crystal data clearly demonstrates a Ca2+ binding site near the NL/NRX binding interface, by direct interactions with residues from the second LNS  domain of NRX, and supported by the interaction with NL (Araç et al., 2007; Chen et al., 2008a; Fabrichny et al., 2007). This binding sites is homologous to the low-affinity site (Kd approximately 400 microM ) previously observed which ceases to be measurable when an 8- or 15-residue splice insert  32 is present at the splice site 2 (Sheckler et al., 2006), indicating that alternative splicing can affect Ca2+-binding sites of neurexin LNS/LG domains. These findings provide insights into the structure-function relationship of neuroligin-neurexin mediated synapse maturation.    Figure 1.13: Structure of the NL1/Nrx1ȕ complex. (A) Ribbon diagram of the NL1/Nrx1ȕ heterotetramer. (B) Overall view of a NL1/Nrx1ȕ heterodimer showing carbohydrates (yellow sticks), splice sites NRX SS4 and NL SSB (arrows) and Ca2+ ions at the binding interface (green spheres). Reproduced with permission from (Araç et al., 2007).   1.5.4 Spatial and temporal distribution  In situ hybridization in adult rat tissue uncovered the presence of NL1, 2, and 3 transcripts in the brain (Ichtchenko et al., 1995; Ichtchenko et al., 1996; Paraoanu et al., 2006). During development, NL1 expression is low before birth, increases the first postnatal week, and remains relatively high into adulthood (Song et al., 1999). The increase of NL1 expression during early postnatal development coincides with the increased level of expression of other synaptic proteins such as PSD-95, during a period of active synaptogenesis (Petralia et al., 2005; Song et al., 1999). NL1 is highly polarized to the dendritic plasma membrane and dendritic targeting relies on a cytoplasmic amino acid motif (Rosales et al., 2005). Immuno-localization studies suggested that different  33 neuroligin isoforms are preferentially targeted: NL1 is primarily localized to the postsynaptic side of glutamatergic synapses, whereas NL2 is preferentially targeted to GABAergic synapses (Figure 1.14)(Chih et al., 2005; Levinson et al., 2005; Song et al., 1999; Varoqueaux et al., 2004).   Figure 1.14: Endogenous localization of neuroligin-1 and -2 in cultured hippocampal neurons. Hippocampal neurons were immunostained for endogenous NL1 or NL2 and either VGLUT or VGAT. (A) At DIV14, both NL1 and 2 strongly cluster at synaptic sites. NL1 clusters are enriched at VGLUT sites (black arrowheads), whereas NL2 clusters are enriched at VGAT sites (white arrowheads). (B) NL1 is weakly detected at VGAT-positive contact sites (white arrowheads), and NL2 is weakly clustered at VGLUT contacts (black arrowheads). Scale bars, 10µm , 1µm. Reproduced from (Levinson et al., 2005).  In developing mice and rats, NL3 is expressed by a variety of glial cells, including immature astrocytes, Schwann cells, satellite glia and olfactory ensheathing glia (Gilbert et al., 2001). NL3 protein levels increase during postnatal development, coinciding with peak synaptogenesis, and are expressed at both glutamatergic and GABAergic synapses. Clustering of NL3 in hippocampal neurons by NRX-expressing cells resulted in accumulation with glutamatergic and GABAergic scaffolding proteins, and co- immunoprecipitation studies revealed the presence of NL1/NL3 and NL2/NL3 complexes in brain extracts. These findings suggest that NL3 is a synaptic adhesion molecule that is  34 a shared component of glutamatergic and GABAergic synapses (Budreck and Scheiffele, 2007).  In contrast, NRX transcripts are detected as early as embryonic day 10 (E10), suggesting a role in developmental processes unrelated to synaptogenesis (Puschel and Betz, 1995). In chick, NRX1 staining revealed an early appearance of this transcript in the retina (E6 and E9), suggesting that NRX1 might have a role in retinal differentiation (Paraoanu et al., 2006). Developmental increase in the ratio of -S4 to +S4 forms of neurexins correlating with an increase in glutamate relative to GABA synaptogenesis and activity regulation of splicing at S4 (Kang et al., 2008).   1.5.5 Synaptic function  NLs were first identified for their ability to bind all three isoforms of β-NRX, a presynaptic transmembrane protein (Ichtchenko et al., 1995). Insights into a direct role for adhesion molecules in synapse development came from an assay involving co-culture of non-neuronal and neuronal cells to study the minimum molecular requirements for synapse induction (Biederer and Scheiffele, 2007). In these experiments, presentation of NL1 or NL2 by heterologous cells was sufficient to induce functional presynaptic terminals in axons contacting these cells, as measured by the accumulation of presynaptic marker such as synapsin, and by ability of these terminals to  could undergo exocytosis in a depolarization-dependent manner (Scheiffele et al., 2000). This feature was later demonstrated for other adhesion molecules such as SynCAM  (Biederer et al., 2002), NGL (Kim et al., 2006), and EphB2 (Kayser et al., 2006). Electrophysiological recordings of artificial contacts formed between neurons and heterologous cells co- expressing NL1 with NMDA or AMPA receptor subunits provided further evidence that neuroligins could induce the formation of functional synapses (Fu et al., 2003). Both spontaneous and action potential-evoked GABAergic events were readily detected in HEK 293T cells co-expressing GABAA receptors with NL2, but not with NL1 (Dong et  35 al., 2007). Conversely, β-NRX presented to dendrites via heterologous cells or beads has been demonstrated to recruit postsynaptic proteins (Graf et al., 2004; Nam and Chen, 2005). In COS cell neuron co-culture assays, all three α-NRXs induce clustering of the GABAergic postsynaptic scaffolding protein gephyrin and NL2 but not of the glutamatergic postsynaptic scaffolding protein PSD-95 or NL 1/3/4 (Kang et al., 2008).   Consistent with the synaptogenic activity of neuroligins in co-culture assays, in neurons the interaction between NL and NRX has been demonstrated to increase the size and number of glutamatergic and GABAergic presynaptic terminals (Figure 1.15)(Levinson et al., 2005; Prange et al., 2004), as well as potentiate the clustering of the postsynaptic protein PSD-95 under some circumstances (Scheiffele et al., 2000), as well as nicotinic synapses (Conroy et al., 2007). These studies were performed using NL that contained both A and B spliced inserts, and is expected to bind only to β-NRX. Over-expression of NL1 lacking the insert at splice site B, which also binds to α-NRX, primarily affected the size of spines and presynaptic terminals, and had less of an effect on the number of synaptic contacts formed (Boucard et al., 2005). Deletion of NL1 in knockout mice selectively decreases the NMDAR/AMPAR ratio, whereas deletion of NL2 selectively decreases inhibitory synaptic responses (Chubykin et al., 2007). Using immunoelectron microscopy, endogenous neurexins and epitope-tagged β-NRX1 are localized to axons and presynaptic terminals in vivo. Unexpectedly, neurexins are also abundant in the postsynaptic density. This cis-expression of β-NRX1 with NL1 was able to inhibit trans- binding to recombinant neurexins, thus blocking the synaptogenic activity of NL1, and reduces the density of presynaptic terminals in cultured hippocampal neurons (Taniguchi et al., 2007). The postsynaptic complex of scaffolding protein PSD-95 and NL can modulate the release probability of transmitter vesicles at synapse in a retrograde way, resulting in altered presynaptic short-term plasticity. Presynaptic beta-neurexin serves as a likely presynaptic mediator of this effect (Futai et al., 2007).    36  Figure 1.15: Neuroligin-1 induces presynaptic contact formation. Hippocampal neurons were transfected HA-NLG or GFP and stained for synaptophysin (Syn). (A) Size and number of Syn (white arrowheads) were enhanced on neurons expressing HA-NLG when compared with GFP-expressing neurons (B). Scale bar 10µm, 1µm. Reproduced from (Prange et al., 2004).   1.5.6 In vivo function: Knock-out strategies  A test for the functional significance of neurexins and neuroligins is to establish whether these proteins are necessary for normal nervous system function in vivo. In drosophila, NRX1 is expressed in central nervous system and highly enriched in synaptic regions of the ventral ganglion and brain. Neurexin-1 null mutants are viable and fertile, but have shortened lifespan. The synapse number is decreased in central nervous system in Neurexin-1 null mutants, and exhibit associative learning defect in larvae (Zeng et al., 2007). Mice null for one NRX -isoform are viable and fertile, whereas mice depleted of more than one isoform die perinatally of respiratory problems. Mice lacking all three - NRX1 show drastically reduced neurotransmitter release at excitatory and inhibitory synapses and die early postnatal (Irina et al., 2007). Detailed morphological analysis of the brains from surviving double knockout mice lacking two -NRXs showed that despite their impaired neurotransmission, there was no gross anatomical defects or changes in the distribution of synaptic proteins, suggesting that -NRXs are not essential for the formation of the vast majority of CNS synapses in vivo but rather regulate the function of these synapses. -NRXs are also required for efficient neurotransmitter release at neuromuscular junctions (Sons et al., 2006). At the NMJ, null mutations prevents normal proliferation of synaptic boutons, causes detachments between pre- and postsynaptic  37 membranes, abnormally long active zones, and result in corresponding alterations in synaptic transmission with reduced quantal content (Li et al., 2007). Although individual NL knockout mice survive and are fertile, the Nlgn1; Nlgn2; Nlgn3 triple knockout mice die shortly after birth owing to respiratory failure (Varoqueaux et al. 2006). Synapses seemed to be morphologically normal in these mice, but the failure rate of evoked transmission was more than ten fold greater than normal at GABAergic/glycinergic synapses, but unchanged at glutamatergic synapses. Deletion of NL1 in knockout mice selectively decreases the NMDAR/AMPAR ratio, whereas deletion of NL2 selectively decreases inhibitory synaptic responses (Chubykin et al. 2007). Taken together, these data indicate that NLs are dispensable for the establishment of synapses, and are important for strengthening synapses via an activity-dependent mechanism, with different NLs acting on distinct types of synapses.  1.6 INHIBITORY SYNAPSE FORMATION  The main inhibitory neurotransmission in the vertebrate CNS is γ-aminobutyric acid (GABA), and is responsible for modulation of every aspect of brain function. The action of GABA is mediated by ionotropic (GABAA) and metabotropic (GABAB) receptors, which are ubiquitously expressed in the CNS (Barnard et al., 1998; Bowery et al., 2002). GABAergic function is fine-tuned at multiple levels, including transmitter synthesis by two isoforms of glutamic acid decarboxylase (GAD); vesicular storage; Ca2+ dependent and independent release; re-uptake in neurons and glial cells; and activation of receptors localized pre-, post-, and extrasynaptically (Kneussel and Betz, 2000b). An ensemble of hundreds of proteins has been identified at the PSD of excitatory synapses, based on proteomics analysis of biochemically purified synaptic preparations (Peng et al., 2004; Walikonis et al., 2000). However, postsynaptic fractions of inhibitory synapses cannot be isolated selectively in biochemical preparations, and accordingly, much less is known about their molecular constituents. The available data indicate that the main difference is  38 the apparent absence of proteins with PDZ domains mediating protein-protein interactions (Sheng and Sala, 2001).  The limited information available on the constituents of inhibitory contacts have hindered our understanding of the mechanisms that govern inhibitory synapse formation and maturation, including the identification of cell adhesion molecules involved. GABAergic synapses are detected earlier than glutamatergic synapses during embryonic brain development and in dissociated neuron cultures (Tyzio et al., 1999; Khazipov et al., 2001; Hennou et al., 2002; (Deng et al., 2007). Most studies on GABAergic synapse formation have focused on the clustering of GABAA receptors and gephyrin, a putative postsynaptic scaffold protein (Dumoulin et al., 2000; Moss and Smart, 2001 Christie et al., 2002b; Danglot et al., 2003; Luscher and Keller, 2004; Studler et al., 2005).  1.6.1 Receptors GABAA receptor subunits can be divided into seven families with multiple isoforms (W hiting et al., 1999). Localization studies using in situ hybridization have shown a broad spatial and temporal heterogeneity of sub-unit expression within the brain (Fritschy and Mohler, 1995; Laurie et al., 1992). In vivo, most neurons contain multiple GABAA receptor subtypes, as shown for hippocampal pyramidal cells, in which α1-, α2- and α5- GABAA receptors are found with a synapse-specific distribution: the α2-subunit is enriched in synapses on the axon initial segment, whereas the α1-subunit is mainly located in somatodendritic synapses and the α5-subunit is extra-synaptic ( Fritschy and Mohler, 1995; Nusser et al., 1996; Crestani et al., 2002). The cell-specific expression and subcellular localization of GABAA receptor subtypes suggests that receptor targeting is unlikely to be determined by presynaptic innervation alone, and protein±protein interaction mechanisms dependent on appropriate sequence motifs present in particular GABAA receptor subunits are likely to ensure appropriate sorting and synaptic targeting. For instance, specific accumulation of GABA(A) receptor subtypes containing α2 subunits at inhibitory synapses is dependent on their ability to bind gephyrin (Tretter et al., 2008). Interestingly, just like at the NMJ there is a measure of pre-patterning:  39 GABAA receptors form clusters on cell membranes even before synapse formation and often smaller than those induced by GABAergic nerve terminals (Brunig et al., 2002; Christie et al., 2002a; Danglot et al., 2003; Studler et al., 2005). A combination of several parallel systems can alter the number of GABA receptors at postsynaptic sites: (1) trapping and release of membrane surface diffusing receptors, also known as horizontal trafficking; (2) endocytosis-mediated internalization and recycling, also known as fast vertical trafficking; and (3) delivery/removal of receptors between the cell surface and organelles for synthesis/degradation, also known as slow vertical trafficking. Delivery of GABA receptors to the surface is thought to involve a combination of motors, adaptor proteins and vesicular cargo (Smith et al., 2006). Like in glutamatergic synapses, several adaptor proteins are also believed to play a dual role as scaffolding complexes. This has lead to the hypothesis that receptor and scaffold adaptor might enter the membrane surface together, and subsequently be subject to membrane diffusion to finally integrate into synaptic sites.  1.6.2 Scaffolding proteins  Gephyrin is a cytoplasmic protein that accumulates at the postsynaptic complex of GABAergic and glycinergic synapses where it forms sub-membranous lattices associated with postsynaptic clusters of receptors (Kneussel and Betz, 2000b). Gephyrin consists of three domains: (1) an N-terminal domain with similarity to the Escherichia coli MogA protein which forms trimers; (2) a central linking region; (3) a C-terminal domain with similarity to E. coli MoeA protein which forms dimers. Both N-terminal trimerization and C-terminal dimerization have been suggested to be involved in gephyrin scaffold formation (Sola et al., 2004; Sola et al., 2001). In addition to its ability to bind receptors, gephyrin has been reported to interact with several other proteins (Figure 1.16). Collybistin, a guanylate exchange factor for cdc-42, interacts with the c-terminal domain and has been shown to be important for postsynaptic localization (Harvey et al., 2004). Gephyrin also interacts with tubulin, (Harvey et al., 2004; Kirsch et al., 1991), as well as  40 key regulators of microfilament dynamics, including profilin, and with microfilament adaptors such as Mena (Giesemann et al., 2003). Finally, gephyrin binding to Dlc (dynein light chain) and other components of myosin motor complexes (Fuhrmann et al., 2002). Thus gephyrin serves as a link between the postsynaptic sites of inhibitory synapses and the cytoskeleton.  Figure 1.16: Postsynaptic gephyrin and its binding partners. Gephyrin is as a sub-membranous scaffold that binds to structural and functional components of inhibitory postsynapse. Gephyrin binds directly to the glycine receptor (GlyR) ȕ subunit and tubulin, is essential for the clustering of a major population of GABAA receptors, and interacts with the tubulin-binding protein GABARAP (GABAA-receptor-associated protein\. The membrane apposition of gephyrin involves binding to PtdIns(3,4,5)P3-anchored proteins implicated in the regulation of actin dynamics and downstream signaling, such as collybistin and profilin. In addition, gephyrin interacts with RAFT1, a candidate regulator of dendritic protein synthesis. Dystrophin, which binds F-actin, might also be crucial for GABAA- receptor clustering. Reproduced with permission from (Kneussel and Betz, 2000a).  At GABAergic synapses, gephyrin is reported to be important for synaptic clustering of GABAA and glycine receptors at inhibitory synapses in hippocampal neurons (Kneussel et al., 1999; Levi et al., 2004).  W hen other regions of the brain were examined, gephyrin was found to be only partially required for their synaptic clustering in retina and spinal cord, and not required for synaptic clustering of other subunits (Fischer et al., 2000; Kneussel et al., 2001). Studies with a gephyrin-deficient mouse mutant have shown that  41 while gephyrin is essential for the synaptic clustering of glycine receptors (Essrich et al., 1998; Levi et al., 2004), and gephyrin is only essential for the clustering of some GABAARs (Kneussel et al., 2001; Kneussel et al., 1999; Levi et al., 2004). In cultured hippocampal pyramidal cells, knock-down of gephyrin decreased both the number of GABAA receptor clusters at postsynaptic sites. (Yu et al., 2007).  GABAA -receptor-associated protein (GABARAP) binds to the abundant γ2 subunit, and has a microtubule binding motif in its N-terminal domain (Wang et al., 1999; Wang and Olsen, 2000). Despite this interaction with GABAA receptors, GABARAP has not been identified at synaptic sites and is not critical for the synaptic delivery of GABAA receptors (Kittler and Moss, 2001; Kneussel and Betz, 2000b; O'Sullivan et al., 2005). The abundance of GABARAP in the Golgi apparatus and its interaction with NSF suggests involvement in the intracellular transport (Moss and Smart, 2001). Sequence similarities with proteins involved in autophagy suggest that GABARAP may be involved in receptor degradation (Kneussel and Loebrich, 2007). GRIP1c, a novel splice form of GRIP1 that lacks PDZ domains 1-3 of GRIP1, concentrates not only in glutamatergic synapses but also in GABAergic synapses both in cultured neurons and in the intact brain (Charych et al., 2004). Considerably higher density of GRIP1c is found in the presynaptic GABAergic terminals than in the glutamatergic terminals, while the density of GRIP1a/b in the postsynaptic complex is similar in both types of synapses. Both splice variants frequently co-localize with each other in individual GABAergic and glutamatergic synapses (Li et al., 2005). Thus, some forms of GRIP1 might play a more significant role in GABAergic synapses than previously recognized.  1.6.3 Cell adhesion molecules  Until recently, few adhesion molecules have been implicated in the formation of inhibitory synapses, and the popular model involved initiation by GABAA receptor  42 activation and signal transduction through activation of G-proteins and phosphatidylinositol 3-kinase (Kneussel and Betz, 2000b). The few CAM s implicated in inhibitory synaptogenesis include dystroglycan, L1, N-CAM, and protocadherins. NL2 represents the first CAM demonstrated to drive inhibitory synapse formation.  Dystroglycan was the first identified adhesive macromolecule at mature GABA synapses (Kneussel et al., 1999; Levi et al., 2002). Dystroglycan is composed of an extracellular Į- subunit and a transmembrane β-subunit derived by proteolytic cleavage and glycosylation of a single precursor protein, and binds several extracellular matrix molecules, and intracellularly binds to dystrophin and utrophin (Williamson et al., 1997). Importantly, dystroglycan also binds to two cell surface proteins involved in synaptogenesis, agrin and β-NRX (Sugita et al., 2001; Sugiyama et al., 1994).  Other members of the dystrophin-associated complex (DPC) also have been implicated. Dystrophin also clusters with a sub-set of GABAergic synapses, and although a direct link with GABAA receptors has not been demonstrated, the absence of dystrophin in mice leads to a reduction in GABAA receptor clustering, but not gephyrin (Brunig et al., 2002; Knuesel et al., 1999). Despite the potential of dystroglycan as an inhibitory synaptic adhesion protein, studies on neurons obtained from mice mutant for dystroglycan indicate that this protein is not essential, since clustering of many proteins that localize to inhibitory synapses (Levi et al., 2002). Interestingly, dystroglycan and dystrophin is highly expressed in brain regions with high synaptic plasticity and appears later in development (Knuesel et al., 2000), suggesting that the DPC may play a more prominent role in maturation and refinement of inhibitory contacts.  L1-CAM  is expressed by almost all post-mitotic neurons in the CNS at the onset of differentiation, and has been shown to be important in neuronal migration, neurite outgrowth, fasciculation and guidance, survival and myelination in the PNS. Some of their roles in the CNS, particularly at synapses, have recently been elucidated. Synaptic plasticity in hippocampal slices from adult constitutively L1-deficient mice do not have significant changes in long-term potentiation (LTP) (Bliss et al., 2000). These findings were re-capitulated in juvenile mice, and interestingly, perisomatic inhibition was  43 noticeably impaired, as well as a decrease in the number of inhibitory synapses (Saghatelyan et al., 2004). These results demonstrate that L1-CAM  can modulate the function as well as the number of inhibitory synapses, but do not demonstrate how L1- CAM exerts this effect. A clue for one of the functions of L1-CAM comes from study of the member neurofascin186 (Ango et al., 2004). The precise localization of neurofascin186 by the membrane adaptor protein ankyrinG, was responsible for establishment of a gradient at the initial axon segment cerebellar purkinje cells. Disruption of the gradient of neurofascin186 results in disruption of the subcellular localization of GABAergic synapses on the axonal initial segment (Ango et al., 2004). These data suggest that L1 may be important in target recognition of inhibitory synapses to precise locations on a target neuron, and warrants further study in other neuronal circuits.  Protocadherins are important for inhibitory synapse development in the spinal cord since neurons cultured form pcdh-γ  mutant mice show a fewer GABAergic terminals. Electrophysiology analysis also revealed a decrease in the spontaneous and evoked inhibitory currents (Weiner et al., 2005). It will be interesting to see if protocadherins are important for inhibitory synapse development in other areas of the brain.  Neuroligin. The first direct evidence of a CAM driving formation of inhibitory synapses came from Prange et al. (2004), with a surprising result. NL1, a well-characterized glutamatergic synapse CAM, also induced the formation of GABAergic synapses when over-expressed in hippocampal neurons (Prange et al., 2004). Later, it was discovered that this was a property of all members of the NL family and their presynaptic binding partner -NRX (Chih et al., 2005; Graf et al., 2004; Levinson et al., 2005). Endogenously however, NL1 and 3 are enriched at excitatory synapses, whereas NL2 is enriched at inhibitory synapses (Levinson et al., 2005; Song et al., 1999; Varoqueaux et al., 2004). Interestingly, the ability of neuroligins to drive formation of excitatory and inhibitory synapses seems to require interaction with β-NRX, the same presynaptic partner for which it was originally identified. Treatment of cultured neurons with a soluble form of β-NRX blocks the formation of both excitatory and inhibitory synapses  44 (Levinson et al., 2005; Scheiffele et al., 2000). How NLs are able to drive the formation of both excitatory and inhibitory synapses using the same presynaptic partner remains unknown. Secondary presynaptic binding partners or cross-talk with other adhesions systems may account for this phenomenon, but remains to be determined.  1.6.4 Presynaptic mechanisms  The contribution of neurotransmitter release in synapse formation has been studied using two approaches: blockade of neurotransmitter release or their receptors, and the development of autapses. In the latter, culturing hippocampal neurons in isolation only autapses are made, and clear miss-matches are observed with clusters of GABA receptors and gephyrin apposed to glutamatergic terminals in pyramidal cells, and clusters of glutamate receptors and PSD-95 apposed to GABAergic presynaptic terminals in interneurons (Rao et al., 2000). The frequency of these miss-matches is greatly reduced in a regular culture, suggesting a µgeneral¶ synaptogenic factor capable of clustering receptors and scaffolds independent of the neurotransmitter involved that can be later over-ridden by neurotransmitter release. This phenomenon is not unique to hippocampal neurons. In cultured cerebellar granule cells, GABAergic terminals are normally surrounded by multiple gephyrin clusters. However, in dendrites devoid of innervation, these gephyrin clusters are distributed randomly, but often apposed to glutamatergic terminals (Studler et al., 2002). In line with the results form these miss-matched synapses, chronic blockade of action potential-mediated neurotransmitter release with tetrodotoxin (TTX), or chronic blockade of GABAA receptors with bicuculline, does not affect their post-synaptic clustering and co-localization with gephyrin (Meier et al., 2001). Thus while communication across the synapse involves multiple signals, activation of GABA receptors appears to be dispensable for building these synapses.     45 1.7 SYNAPTIC IMBALANCE IN AUTISM  Several molecules co-ordinate formation of excitatory and inhibitory synapses, and dysfunction of these molecular systems are believed to be important in the etiology of developmental psychiatric disorders, such as autism, where abnormal neuronal wiring and imbalance in excitatory/inhibitory (E/I) ratio has occured (Cline, 2005).  1.7.1 Definition and etiology  Autism spectrum disorders (ASDs) are characterized by complex behavioral and cognitive deficits including abnormal social interaction and communication, repetitive behavior and atypical information processing (American Psychiatric Association: 1994). A number of studies have confirmed that ASDs has an important genetic component, and is due to a disruption in experience-dependent synaptic activity and an imbalance between E/I synaptic transmission (Wassink et al., 2004). Enhanced E/I neurotransmission due to either enhanced excitation or reduced inhibition, can result in the defects in social and behavioral abnormalities associated with this disorder. Thus, molecules important in the establishment and function of synapses are expected to play a role in the expression of ASDs.  1.7.2 Genetic links  Several genes important for synapse formation show chromosomal aberrations in autistic patients (Table 1.1).  The ability of NLs to form both excitatory and inhibitory synapses, makes them attractive candidates for the etiology of ASD. Consistent with this, rearrangement of chromosomal regions containing NL genes and mutations in the genes themselves, have been linked to ASD (Jamain et al., 2003; Laumonnier et al., 2004; Yan et al., 2004).  Other studies have indicated that NL mutations are not common among all ASD patients (Anne-Kathrin Wermter, 2008; Vincent et al., 2004; Ylisaukko-oja et al.,  46 2005), and is not surprising considering the heterogeneity of ASD expression, It is likely that additional candidates within the synaptic protein families will be implicated.  In vitro studies have provided further understanding of how mutations in NLs associated with ASD might alter cellular mechanisms and lead to synapse dysfunction. A defect in cell surface transport of some of NL mutants was detected and correlated with an increase in accumulation in the endoplasmic reticulum, presumably due to protein miss-folding (Chih et al., 2004; Chubykin et al., 2005). When expressed in hippocampal neurons, these mutants showed signs of defective function, as measured by a loss of ability to promote presynaptic differentiation and reduced affinity for NRX (Chih et al., 2004; Comoletti et al., 2004). A small percentage of mutants were still trafficked to the surface and their synapse-inducing ability was preserved, indicating that ASD-associated mutations do not completely eliminate synaptogenic activity of NLs, but may cause other changes in synapse physiology. Indeed, introduction of a point mutation in NL1 similar to the NL3 mutation (R451C) associated with ASD, can increase NMDA receptor clustering and current amplitude (Khosravani et al., 2005). Thus, phenotypes caused by synaptic protein mutations are likely to be important for explaining some of the synaptic defects associated with ASD.   47 Table 1.2: Chromosome abnormalities associated with autism Protein Chromosom e Linkage Ref Neuroligin 1 3q26.3 ~Chromosomal abnormality ~Linkage in family members (Auranen et al., 2002; 2003; Konstantareas and Homatidis, 1999) Neuroligin 2 17p13.2 ~Chromosomal abnormalities ~Linkage in affected siblings (Mariner et al., 1986; Risch et al., 1999) Neuroligin 3 Xq13.1 ~Point mutation (R451C). (Jamain et al., 2003) Neuroligin 4 Xp22.33 ~De novo chromosomal deletions in three males with autism. ~Frame shift mutation (D396X) ~2-base-pair deletion found (1253delAG) ~Four miss-sense mutations in unrelated patients (G99S,F378R,V403M, R704C). (Jamain et al., 2003; Laumonnier et al., 2004; Lawson et al., 2008; Marshall et al., 2008; Thomas et al., 1999; Yan et al., 2004) Neurexin 1 2p16.3 ~Alpha and beta affected ~incomplete penetrence (Feng et al., 2006; Kim et al., 2008; Marshall et al.,2008) PSD-95 17p13.2 ~ Linkage in siblings affected with autism (Risch et al., 1999) GABARAP 17p13.2 ~ Linkage in siblings affected with autism. (Risch et al., 1999) Shank3 22q13.3 ~ Linkage in siblings affected with autism. (Durand et al., 2007; Marshall et al., 2008; Moessner et al., 2007)  1.7.3 Mouse models Expression of the Arg451-->Cys451 (R451C) substitution in NL3 mutant mice showed impaired social interactions but enhanced spatial learning abilities. Unexpectedly, these behavioral changes were accompanied by an increase in inhibitory synaptic transmission with no apparent effect on excitatory synapses. Deletion of NL3, in contrast, did not cause such changes, indicating that the R451C substitution represents a gain-of-function mutation. These data suggest that increased inhibitory synaptic transmission may contribute to human ASDs and that the R451C knock-in mice may be a useful model for studying autism-related behaviors.(Irina Dudanova, 2007). Mutant mice lacking the murine ortholog NL4 exhibit highly selective deficits in reciprocal social interactions and communication reminiscent of autistic spectrum disorders in humans. Thus, NL-4-KOs represent a genetic animal model of non-syndromic monogenic heritable ASDs (Jamain et al., 2008).  48 1.8 RATIONALE AND HYPOTHESIS  The postsynaptic density protein PSD-95 is critical for assembly and clustering of many molecules, including scaffold proteins, cell adhesion molecules, and ion channels. Studies on cultured hippocampal neurons showed that PSD-95 and another postsynaptic protein GKAP start to accumulate at synapses within the first few days after plating (Rao et al., 1998). This accumulation of PSD-95 and GKAP proceeded clustering of other synaptic proteins, including glutamate receptors. Due to its vital role in orchestrating the assembly of several molecules critical for synapse formation and maturation, I hypothesize that PSD-95 induces the formation of a pre-assembled protein complex containing core molecules important for nucleation of the postsynaptic density. The NL family of proteins, and their binding partners the neurexins, have been shown to be important in establishing both glutamatergic and GABAergic synapses (Biederer, 2005; Cline, 2005; Dean and Dresbach, 2006; Gilbert and Auld, 2005; Levinson and El- Husseini, 2005a; Levinson and El-Husseini, 2005b). Endogenously NL1 is mainly restricted to glutamatergic synapses and NL2 is mainly restricted to GABAergic synapse (Levinson et al., 2005; Varoqueaux et al., 2004). However, over-expression of scaffolding molecules such as PSD-95 can manipulate the clustering and the number of synapses induced by NL1. Conversely, preliminary data show that scaffolding molecules such as gephyrin, which localized at inhibitory synapses, can modulate clustering of NL2 at GABAergic contacts. Based on these findings, it appears that NLs have the ability to affect both excitatory and inhibitory synapse development, but it is their interaction with scaffolding molecules that determines their site of action. I hypothesize that scaffolding proteins may modulate the differential sorting of NLs by regulating their sequestration and/or retention to neuronal contacts to regulate excitatory and inhibitory contact development. The primary goal of this work is to clarify mechanisms that control differential sorting, trafficking and clustering of NLs and their intacellular binding partners at neuronal contacts. Together these results will give insight into the mechanisms of protein trafficking during synapse formation, and provide a better understanding of the etiology of developmental diseases characterized by synaptic dysfunction.  49 1.9 REFERENCES  Ahmari, S. E., Buchanan, J., and Smith, S. J. (2000). Assembly of presynaptic active zones from cytoplasmic transport packets. Nat Neurosci 3, 445-451.  Ahmari, S. E., and Smith, S. J. (2002). Knowing a nascent synapse when you see it. Neuron 34, 333-336.  Anderson, M. J., Cohen, M.W. (1977). Nerve-induced and spontaneous redistribution of acetylcholine receptors on cultured muscle cells. J Physiol 268, 757-773.  Ango, F., di Cristo, G., Higashiyama, H., Bennett, V., Wu, P., and Huang, Z. J. (2004). Ankyrin-based subcellular gradient of neurofascin, an immunoglobulin family protein, directs GABAergic innervation at purkinje axon initial segment. Cell 119, 257-272.  Anne-Kathrin Wermter, I. K.-B., Konstantin Strauch, Gerd Schulte-Körne, Helmut Remschmidt, (2008). No evidence for involvement of genetic variants in the X-linked neuroligin genes NLGN3 and NLGN4X in probands with autism spectrum disorder on high functioning level. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 9999  Aoki, C., Miko, I., Oviedo, H., Mikeladze-Dvali, T., Alexandre, L., Sweeney, N., and Bredt, D. S. (2001). Electron microscopic immunocytochemical detection of PSD-95, PSD-93, SAP-102, and SAP-97 at postsynaptic, presynaptic, and nonsynaptic sites of adult and neonatal rat visual cortex. Synapse 40, 239- 257.  Araç, D., Boucard, A. A., Ozkan, E., Strop, P., Newell, E., Südhof, T. C., and Brunger, A. T. (2007). Structures of neuroligin-1 and the neuroligin-1/neurexin-1 beta complex reveal specific protein-protein and protein-Ca2+ interactions. Neuron 56, 992-1003.  Arber, S., Burden, S.J., Harris, A.J. (2002). Patterning of skeletal muscle. Curr Opin Neurobiol 12, 100- 103.  Auranen, M., Vanhala, R., Varilo, T., Ayers, K., Kempas, E., Ylisaukko-Oja, T., Sinsheimer, J. S., Peltonen, L., and Jarvela, I. (2002). A genomewide screen for autism-spectrum disorders: evidence for a major susceptibility locus on chromosome 3q25-27. Am J Hum Genet 71, 777-790.  Auranen, M., Varilo, T., Alen, R., Vanhala, R., Ayers, K., Kempas, E., Ylisaukko-oja, T., Peltonen, L., and Jarvela, I. (2003). Evidence for allelic association on chromosome 3q25-27 in families with autism spectrum disorders originating from a subisolate of Finland. 8, 879-884.  Balice-Gordon, R. J., and Lichtman, J. W. (1994). Long-term synapse loss induced by focal blockade of postsynaptic receptors. Nature 372, 519-524.  Bamji, S. X., Shimazu, K., Kimes, N., Huelsken, J., Birchmeier, W., Lu, B., and Reichardt, L. F. (2003). Role of beta-catenin in synaptic vesicle localization and presynaptic assembly. Neuron 40, 719-731.  Barnard, E. A., Skolnick, P., Olsen, R. W., Mohler, H., Sieghart, W., Biggio, G., Braestrup, C., Bateson, A. N., and Langer, S. Z. (1998). International Union of Pharmacology. XV. Subtypes of gamma-aminobutyric acidA receptors: classification on the basis of subunit structure and receptor function. Pharmacol Rev 50, 291-313.  Bats, C., Groc, L., Choquet, D. (2007). The interaction between Stargazin and PSD-95 regulates AMPA receptor surface trafficking. Neuron 53, 719-34.   50 Bennett, M. R., and Robinson, J. (1989). Growth and elimination of nerve terminals at synaptic sites during polyneuronal innervation of muscle cells: a trophic hypothesis. Proc R Soc Lond B Biol Sci Jan 23; 235, 299-320.  Benson, D. L., and Tanaka, H. (1998). N-cadherin redistribution during synaptogenesis in hippocampal neurons. J Neurosci 18, 6892-6904.  Biederer, T. (2005). Progress from the postsynaptic side: signaling in synaptic differentiation. Sci STKE 2005, pe9.  Biederer, T., Sara, Y., Mozhayeva, M., Atasoy, D., Liu, X., Kavalali, E. T., and Sudhof, T. C. (2002). SynCAM, a synaptic adhesion molecule that drives synapse assembly. Science 297, 1525-1531.  Biederer, T., and Scheiffele, P. (2007). Mixed-culture assays for analyzing neuronal synapse formation. 2, 670-676.  Biederer, T., and Sudhof, T. C. (2001). CASK and Protein 4.1 Support F-actin Nucleation on Neurexins. J Biol Chem 276, 47869-47632.  Bladt, F., Tafuri, A., Gelkop, S., Langille, L., and Pawson, T. (2002). Epidermolysis bullosa and embryonic lethality in mice lacking the multi-PDZ domain protein GRIP1. Proc Natl Acad Sci U S A 99, 6816-6821.  Bliss, T., Errington, M., Fransen, E., Godfraind, J. M., Kauer, J. A., Kooy, R. F., Maness, P. F., and Furley, A. J. (2000). Long-term potentiation in mice lacking the neural cell adhesion molecule L1. Curr Biol 10, 1607-1610.  Boeckers, T. (2006). The postsynaptic density. Cell and Tissue Research 326, 409-422.  Bolliger, M. F., Frei, K., Winterhalter, K. H., and Gloor, S. M. (2001 ). Identification of a novel neuroligin in humans which binds to PSD-95 and has a widespread expression. . Biochem J 356, 581-588.  Borgdorff, A. J., and Choquet, D. (2002). Regulation of AMPA receptor lateral movements. 417, 649-653.  Bormann, J. (2000). The 'ABC' of GABA receptors. Trends Phamacol Sci 21, 16-19.  Boucard, A. A., C., Comoletti, D., Taylor, P., and Sudhof, T. C. (2005). A splice code for trans-synaptic cell adhesion mediated by binding of neuroligin 1 to alpha- and beta-neurexins. Neuon 48, 229-236.  Bowery, N. G., Bettler, B., Froestl, W., Gallagher, J. P., Marshall, F., Raiteri, M., Bonner, T. I., and Enna, S. J. (2002). International Union of Pharmacology. XXXIII. Mammalian gamma-aminobutyric acid(B) receptors: structure and function. Pharmacol Rev 54, 247-264.  Bozdagi, O., Valcin, M., Poskanzer, K., Tanaka, H., and Benson, D. L. (2004). Temporally distinct demands for classic cadherins in synapse formation and maturation. Mol Cell Neurosci 27, 509-521.  Bredt, D. S., and Nicoll, R. A. (2003). AMPA receptor trafficking at excitatory synapses. Neuron 40, 361- 379.  Bresler, T., Ramati, Y., Zamorano, P. L., Zhai, R., Garner, C. C., and Ziv, N. E. (2001). The dynamics of SAP90/PSD-95 recruitment to new synaptic junctions. Mol Cell Neurosci 18, 149-167.  Bresler, T., Shapira, M., Boeckers, T., Dresbach, T., Futter, M., Garner, C. C., Rosenblum, K., Gundelfinger, E. D., and Ziv, N. E. (2004). Postsynaptic density assembly is fundamentally different from presynaptic active zone assembly. J Neurosci 24, 1507-1520.   51 Brose, N. (1999). Synaptic cell adhesion proteins and synaptogenesis in the mammalian central nervous system. Naturwissenschaften 86, 516-524.  Brunig, I., Suter, A., Knuesel, I., Luscher, B., and Fritschy, J. M. (2002). GABAergic terminals are required for postsynaptic clustering of dystrophin but not of GABA(A) receptors and gephyrin. J Neurosci 22, 4805-4813.  Budreck, E. C., and Scheiffele, P. (2007). Neuroligin-3 is a neuronal adhesion protein at GABAergic and glutamatergic synapses. European Journal of Neuroscience 26, 1738-1748.  Butz, S., Okamoto, M., and Sudhof, T. C. (1998). A tripartite protein complex with the potential to couple synaptic vesicle exocytosis to cell adhesion in brain. Cell 94, 773-782.  Carlisle, H. J., and Kennedy, M. B. (2005). Spine architecture and synaptic plasticity. Trends Neurosci 28, 182-187.  Charych, E. I., Yu, W., Li, R., Serwanski, D. R., Miralles, C. P., Li, X., Yang, B. Y., Pinal, N., Walikonis, R., and De Blas, A. L. (2004). A four PDZ domain-containing splice variant form of GRIP1 is localized in GABAergic and glutamatergic synapses in the brain. J Biol Chem 279, 38978-38990.  Chavis, P., and Westbrook, G. (2001). Integrins mediate functional pre- and postsynaptic maturation at a hippocampal synapse. Nature 411, 317-321.  Chen, L., Chetkovich, D. M., Petralia, R. S., Sweeney, N. T., Kawasaki, Y., Wenthold, R. J., Bredt, D. S., and Nicoll, R. A. (2000). Stargazing regulates synaptic targeting of AMPA receptors by two distinct mechanisms. Nature 408, 936-943.  Chen, X., Liu, H., Shim, A. H. R., Focia, P. J., and He, X. (2008a). Structural basis for synaptic adhesion mediated by neuroligin-neurexin interactions. 15, 50-56.  Chen, X., Winters, C., Azzam, R., Li, X., Galbraith, J. A., Leapman, R. D., and Reese, T. S. (2008b). Organization of the core structure of the postsynaptic density. Proceedings of the National Academy of Sciences 105, 4453-4458.  Chih, B., Afridi, S. K., Clark, L., and Scheiffele, P. (2004). Disorder-associated mutations lead to functional inactivation of neuroligins. Hum Mol Genet.  Chih, B., Engelman, H., and Scheiffele, P. (2005). Control of Excitatory and Inhibitory Synapse Formation by Neuroligins. Science 307, 1324-1328.  Chubykin, A. A., Atasoy, D., Etherton, M. R., Brose, N., Kavalali, E. T., Gibson, J. R., and Südhof, T. C. (2007). Activity-Dependent Validation of Excitatory versus Inhibitory Synapses by Neuroligin-1 versus Neuroligin-2. Neuron 54, 919-931.  Chubykin, A. A., Liu, X., Comoletti, D., Tsigelny, I., Taylor, P., and Sudhof, T. C. (2005). Dissection of synapse induction by neuroligins: Effect of a neuroligin mutation associated with autism. J Biol Chem.  Cline, H. (2005). Synaptogenesis: a balancing act between excitation and inhibition. Curr Biol 15, R203- 205.  Cline, H. T. (2001). Dendritic arbor development and synaptogenesis. Curr Opin Neurobiol 11, 118-126.  Comoletti, D., De Jaco, A., Jennings, L. L., Flynn, R. E., Gaietta, G., Tsigelny, I., Ellisman, M. H., and Taylor, P. (2004). The Arg451Cys-neuroligin-3 mutation associated with autism reveals a defect in protein processing. J Neurosci 24, 4889-4893.   52 Comoletti, D., Flynn, R., Jennings, L. L., Chubykin, A., Matsumura, T., Hasegawa, H., Sudhof, T. C., and Taylor, P. (2003). Characterization of the Interaction of a Recombinant Soluble Neuroligin-1 with Neurexin-1{beta}. J Biol Chem 278, 50497-50505.  Comoletti, D., Grishaev, A., Whitten, A. E., Tsigelny, I., Taylor, P., and Trewhella, J. (2007). Synaptic arrangement of the neuroligin/beta-neurexin complex revealed by X-ray and neutron scattering. Structure 15, 693-705.  Conroy, W. G., Nai, Q., Ross, B., Naughton, G., and Berg, D. K. (2007). Postsynaptic neuroligin enhances presynaptic inputs at neuronal nicotinic synapses. Developmental Biology 307, 79-91.  Craig, A. M., and Boudin, H. (2001). Molecular heterogeneity of central synapses: afferent and target regulation. Nat Neurosci 4, 569-578.  Craig, A. M., Graf, E. R., and Linhoff, M. W. (2006). How to build a central synapse: clues from cell culture. Trends Neurosci 29, 8-20.  Craig, A. M., and Kang, Y. (2007). Neurexin-neuroligin signaling in synapse development. Current Opinion in NeurobiologyDevelopment 17, 43-52.  Crestani, F., Keist, R., Fritschy, J.-M., Benke, D., Vogt, K., Prut, L., Bluthmann, H., Mohler, H., and Rudolph, U. (2002). Trace fear conditioning involves hippocampal alpha 5 GABAA receptors. Proceedings of the National Academy of Sciences 99, 8980-8985.  Dahan, M., Levi, S., Luccardini, C., Rostaing, P., Riveau, B., and Triller, A. (2003). Diffusion dynamics of glycine receptors revealed by single-quantum dot tracking. Science 302, 442-445.  Dalva, M. B., Takasu, M. A., Lin, M. Z., Shamah, S. M., Hu, L., Gale, N. W., and Greenberg, M. E. (2000). EphB receptors interact with NMDA receptors and regulate excitatory synapse formation. Cell 103, 945-956.  Dean, C., and Dresbach, T. (2006). Neuroligins and neurexins: linking cell adhesion, synapse formation and cognitive function. Trends Neurosci 29, 21-29.  Dean, C., Scholl, F. G., Choih, J., DeMaria, S., Berger, J., Isacoff, E., and Scheiffele, P. (2003). Neurexin mediates the assembly of presynaptic terminals. Nat Neurosci 6, 708-716.  DeChiara, T. M., Bowen, D.C., Valenzuela, D.M., Simmons, M.V., Poueymirou, W.T., Thomas, S., Kinetz, E., Compton, D.L., Rojas, E., Park, J.S., Smith, C., DiStefano, P.S., Glass, D.J., Burden, S.J., Yancopoulos, G.D. (1996). The receptor tyrosine kinase MuSK is required for neuromuscular junction formation in vivo. Cell 85, 501-512.  Deng, F., Price, M. G., Davis, C. F., Mori, M., and Burgess, D. L. (2006). Stargazin and Other Transmembrane AMPA Receptor Regulating Proteins Interact with Synaptic Scaffolding Protein MAGI-2 in Brain. J Neurosci 26, 7875-7884.  Deng, L., Yao, J., Fang, C., Dong, N., Luscher, B., and Chen, G. (2007). Sequential Postsynaptic Maturation Governs the Temporal Order of GABAergic and Glutamatergic Synaptogenesis in Rat Embryonic Cultures. J Neurosci 27, 10860-10869.  DeSouza, S., Fu, J., States, B. A., and Ziff, E. B. (2002). Differential palmitoylation directs the AMPA receptor-binding protein ABP to spines or to intracellular clusters. J Neurosci 22, 3493-3503.  Dong, H., O'Brien, R. J., Fung, E. T., Lanahan, A. A., Worley, P. F., and Huganir, R. L. (1997). GRIP: a synaptic PDZ domain-containing protein that interacts with AMPA receptors. Nature 386, 279-284.   53 Dong, H., Zhang, P., Song, I., Petralia, R. S., Liao, D., and Huganir, R. L. (1999). Characterization of the Glutamate Receptor-Interacting Proteins GRIP1 and GRIP2. J Neurosci 19, 6930-6941.  Dong, N., Qi, J., and Chen, G. (2007). Molecular reconstitution of functional GABAergic synapses with expression of neuroligin-2 and GABAA receptors. Molecular and Cellular Neuroscience 35, 14-23.  Durand, C. M., Betancur, C., Boeckers, T. M., Bockmann, J., Chaste, P., Fauchereau, F., Nygren, G., Rastam, M., Gillberg, I. C., Anckarsater, H., et al. (2007). Mutations in the gene encoding the synaptic scaffolding protein SHANK3 are associated with autism spectrum disorders. 39, 25-27.  Eaton, B. A., Fetter, R. D., and Davis, G. W. (2002). Dynactin is necessary for synapse stabilization. Neuron 34, 729-741.  Ehrlich, I., and Malinow, R. (2004). Postsynaptic density 95 controls AMPA receptor incorporation during long-term potentiation and experience-driven synaptic plasticity. J Neurosci 24, 916-927.  El-Husseini, A. E., Schnell, E., Chetkovich, D. M., Nicoll, R. A., and Bredt, D. S. (2000a). PSD-95 involvement in maturation of excitatory synapses. Science 290, 1364-1368.  El-Husseini, A. E., Topinka, J. R., Lehrer-Graiwer, J. E., Firestein, B. L., Craven, S. E., Aoki, C., and Bredt, D. S. (2000b). Ion channel clustering by membrane-associated guanylate kinases. Differential regulation by N-terminal lipid and metal binding motifs. J Biol Chem 275, 23904-23910.  Essrich, C., Lorez, M., Benson, J. A., Fritschy, J.-M., and Luscher, B. (1998). Postsynaptic clustering of major GABAA receptor subtypes requires the [gamma]2 subunit and gephyrin. 1, 563-571.  Ethell, I. M., Irie, F., Kalo, M. S., Couchman, J. R., Pasquale, E. B., and Yamaguchi, Y. (2001). EphB/syndecan-2 signaling in dendritic spine morphogenesis. Neuron 31, 1001-1013.  Ethell, I. M., and Pasquale, E. B. (2005). Molecular mechanisms of dendritic spine development and remodeling. Prog Neurobiol 75, 161-205.  Fabrichny, I. P., Leone, P., Sulzenbacher, G., Comoletti, D., Miller, M. T., Taylor, P., Bourne, Y., and Marchot, P. (2007). Structural analysis of the synaptic protein neuroligin and its beta-neurexin complex: determinants for folding and cell adhesion. Neuron 56, 979-991.  Feng, J., Schroer, R., Yan, J., Song, W., Yang, C., Bockholt, A., Cook, J., Edwin H., Skinner, C., Schwartz, C. E., and Sommer, S. S. (2006). High frequency of neurexin 1[beta] signal peptide structural variants in patients with autism. Neuroscience Letters 409, 10-13.  Feng, J., Yan, Z., Ferreira, A., Tomizawa, K., Liauw, J. A., Zhuo, M., Allen, P. B., Ouimet, C. C., and Greengard, P. (2000). Spinophilin regulates the formation and function of dendritic spines. Proc Natl Acad Sci U S A 97, 9287-9292.  Ferreira, A., and Paganoni, S. (2002). The formation of synapses in the central nervous system. Mol Neurobiol 26, 69-79.  Fiala, J. C., Feinberg, M., Popov, V., and Harris, K. M. (1998). Synaptogenesis via dendritic filopodia in developing hippocampal area CA1. J Neurosci 18, 8900-8911.  Fischer, F., Kneussel, M., Tintrup, H., Haverkamp, S., Rauen, T., Betz, H., and Wassle, H. (2000). Reduced synaptic clustering of GABA and glycine receptors in the retina of the gephyrin null mutant mouse. J Comp Neurol 427, 634-648.   54 Frank, E., Fischbach, G.D. (1979). Early events in neuromuscular junction formation in vitro: induction of acetylcholine receptor clusters in the postsynaptic membrane and morphology of newly formed synapses. J Cell Biol 83, 143-158.  Friedman, H. V., Bresler, T., Garner, C. C., and Ziv, N. E. (2000). Assembly of new individual excitatory synapses: time course and temporal order of synaptic molecule recruitment. Neuron 27, 57-69.  Fritschy, J. M., and Mohler, H. (1995). GABAA-receptor heterogeneity in the adult rat brain: differential regional and cellular distribution of seven major subunits. J Comp Neurol 359, 154-194.  Fu, Z., Washbourne, P., Ortinski, P., and Vicini, S. (2003). Functional excitatory synapses in HEK293 cells expressing neuroligin and glutamate receptors. J Neurophysiol 90, 3950-3957.  Fuhrmann, J. C., Kins, S., Rostaing, P., El Far, O., Kirsch, J., Sheng, M., Triller, A., Betz, H., and Kneussel, M. (2002). Gephyrin interacts with Dynein light chains 1 and 2, components of motor protein complexes. J Neurosci 22, 5393-5402.  Futai, K., Kim, M. J., Hashikawa, T., Scheiffele, P., Sheng, M., and Hayashi, Y. (2007). Retrograde modulation of presynaptic release probability through signaling mediated by PSD-95-neuroligin. 10, 186- 195.  Garner, C. C., Zhai, R. G., Gundelfinger, E. D., and Ziv, N. E. (2002). Molecular mechanisms of CNS synaptogenesis. Trends Neurosci 25, 243-251.  Gautam, M., Noakes, P. G., Moscoso, L., Rupp, F., Scheller, R. H., Merlie, J. P., and Sanes, J. R. (1996). Defective neuromuscular synaptogenesis in agrin-deficient mutant mice. Cell 85, 525-535.  Gerrow, K., and El-Husseini, A. (2006). Cell adhesion molecules at the synapse. Front Biosci 11, 2400- 2419.  Gerrow, K., and El-Husseini, A. (2007). Receptors look outward: revealing signals that bring excitation to synapses. Sci STKE 2007, pe56.  Gerrow, K., Romorini, S., Nabi, S.,  Colicos, M., Sala, C.; El-Husseini, A. (2005). A preformed complex of postsynaptic proteins is involved in excitatory synapse development. Neuron 49, 547-562.  Giesemann, T., Schwarz, G., Nawrotzki, R., Berhorster, K., Rothkegel, M., Schluter, K., Schrader, N., Schindelin, H., Mendel, R. R., Kirsch, J., and Jockusch, B. M. (2003). Complex formation between the postsynaptic scaffolding protein gephyrin, profilin, and Mena: a possible link to the microfilament system. J Neurosci 23, 8330-8339.  Gilbert, M., Smith, J., Roskams, A. J., and Auld, V. J. (2001). Neuroligin 3 is a vertebrate gliotactin expressed in the olfactory ensheathing glia, a growth-promoting class of macroglia. Glia 34, 151-164.  Gilbert, M. M., and Auld, V. J. (2005). Evolution of clams (cholinesterase-like adhesion molecules): structure and function during development. Front Biosci 10, 2177-2192.  Glass, D. J., Bowen, D. C., Stitt, T. N., Radziejewski, C., Bruno, J., Ryan, T. E., Gies, D. R., Shah, S., Mattsson, K., Burden, S. J., et al. (1996). Agrin acts via a MuSK receptor complex. Cell 85, 513-523.  Goda, Y. (2002). Cadherins communicate structural plasticity of presynaptic and postsynaptic terminals. Neuron 35, 1-3.  Goda, Y., and Davis, G. W. (2003). Mechanisms of synapse assembly and disassembly. Neuron 40, 243- 264.   55 Gorski, J. A., Gomez, L. L., Scott, J. D., and Dell'Acqua, M. L. (2005). Association of an A-Kinase- anchoring Protein Signaling Scaffold with Cadherin Adhesion Molecules in Neurons and Epithelial Cells. Mol Biol Cell 16, 3574-3590.  Graf, E. R., Zhang, X., Jin, S. X., Linhoff, M. W., and Craig, A. M. (2004). Neurexins Induce Differentiation of GABA and Glutamate Postsynaptic Specializations via Neuroligins. Cell 119, 1013- 1026.  Groc, L., Gustafsson, B., and Hanse, E. (2003). In vivo evidence for an activity-independent maturation of AMPA/NMDA signaling in the developing hippocampus. Neuroscience 121, 65-72.  Haas, I. G., Frank, M., Veron, N., and Kemler, R. (2005). Presenilin-dependent processing and nuclear function of gamma-protocadherins. J Biol Chem 280, 9313-9319.  Hambsch, B., Grinevich, V., Seeburg, P. H., and Schwarz, M. K. (2005). {gamma}-Protocadherins, presenilin-mediated release of C-terminal fragment promotes locus expression. J Biol Chem 280, 15888- 15897.  Hannah, M. J., Schmidt, A. A., and Huttner, W. B. (1999). SYNAPTIC VESICLE BIOGENESIS doi:10.1146/annurev.cellbio.15.1.733. Annual Review of Cell and Developmental Biology 15, 733-798.  Harvey, K., Duguid, I. C., Alldred, M. J., Beatty, S. E., Ward, H., Keep, N. H., Lingenfelter, S. E., Pearce, B. R., Lundgren, J., Owen, M. J., et al. (2004). The GDP-GTP exchange factor collybistin: an essential determinant of neuronal gephyrin clustering. J Neurosci 24, 5816-5826.  Hashimoto, K., and Kano, M. (2003). Functional differentiation of multiple climbing fiber inputs during synapse elimination in the developing cerebellum. Neuron 38, 785-796.  Hata, Y., Butz, S., and Sudhof, T. (1996). CASK: a novel dlg/PSD95 homolog with an N-terminal calmodulin-dependent protein kinase domain identified by interaction with neurexins. J Neurosci 16, 2488- 2494.  Hering, H., and Sheng, M. (2001). Dendritic spines: structure, dynamics and regulation. Nat Rev Neurosci 2, 880-888.  Hirao, K., Hata, Y., Ide, N., Takeuchi, M., Irie, M., Yao, I., Deguchi, M., Toyoda, A., Sudhof, T. C., and Takai, Y. (1998). A novel multiple PDZ domain-containing molecule interacting with N-methyl-D- aspartate receptors and neuronal cell adhesion proteins. J Biol Chem 273, 21105-21110.  Hirbec, H., Perestenko, O., Nishimune, A., Meyer, G., Nakanishi, S., Henley, J. M., and Dev, K. K. (2002). The PDZ proteins PICK1, GRIP, and syntenin bind multiple glutamate receptor subtypes. Analysis of PDZ binding motifs. J Biol Chem 277, 15221-15224.  Hoogenraad, C. C., Milstein, A. D., Ethell, I. M., Henkemeyer, M., and Sheng, M. (2005). GRIP1 controls dendrite morphogenesis by regulating EphB receptor trafficking. Nat Neurosci.  Hopf, F. W., Waters, J., Mehta, S., and Smith, S. J. (2002). Stability and plasticity of developing synapses in hippocampal neuronal cultures. J Neurosci 22, 775-781.  Huh, K. H., and Wenthold, R. J. (1999). Turnover analysis of glutamate receptors identifies a rapidly degraded pool of the N-methyl-D-aspartate receptor subunit, NR1, in cultured cerebellar granule cells. J Biol Chem 274, 151-157.  Ichtchenko, K., Hata, Y., Nguyen, T., Ullrich, B., Missler, M., Moomaw, C., and Sudhof, T. C. (1995). Neuroligin 1: a splice site-specific ligand for beta-neurexins. Cell 81, 435-443.   56 Ichtchenko, K., Nguyen, T., and Südhof, T. C. (1996). Structures, Alternative Splicing, and Neurexin Binding of Multiple Neuroligins. J Biol Chem 271, 2676-2682.  Ide, N., Hata, Y., Deguchi, M., Hirao, K., Yao, I., and Takai, Y. (1999). Interaction of S-SCAM with neural plakophilin-related Armadillo-repeat protein/delta-catenin. Biochem Biophys Res Commun 256, 456-461.  Iida, J., Hirabayashi, S., Sato, Y., and Hata, Y. (2004). Synaptic scaffolding molecule is involved in the synaptic clustering of neuroligin. Molecular and Cellular Neuroscience 27, 497-508.  Iida, J., Ishizaki, H., Okamoto-Tanaka, M., Kawata, A., Sumita, K., Ohgake, S., Sato, Y., Yorifuji, H., Nukina, N., Ohashi, K., et al. (2007). Synaptic Scaffolding Molecule {alpha} Is a Scaffold To Mediate N- Methyl-D-Aspartate Receptor-Dependent RhoA Activation in Dendrites. Mol Cell Biol 27, 4388-4405.  Inoue, A., and Sanes, J. R. (1997). Lamina-specific connectivity in the brain: regulation by N-cadherin, neurotrophins, and glycoconjugates. Science 276, 1428-1431.  Irie, M., Hata, Y., Takeuchi, M., Ichtchenko, K., Toyoda, A., Hirao, K., Takai, Y., Rosahl, T. W., and Sudhof, T. C. (1997). Binding of Neuroligins to PSD-95. Science 277, 1511-1515.  Irina Dudanova, K. T., Astrid Rohlmann, Thomas C. Südhof, Markus Missler, (2007). Deletion of alpha- neurexins does not cause a major impairment of axonal pathfinding or synapse formation. The Journal of Comparative Neurology 502, 261-274.  Jackson, H., and Parks, T. N. (1982). Functional synapse elimination in the developing avian cochlear nucleus with simultaneous reduction in cochlear nerve axon branching. J Neurosci 2, 1736-1743.  Jamain, S., Quach, H., Betancur, C., Rastam, M., Colineaux, C., Gillberg, I. C., Soderstrom, H., Giros, B., Leboyer, M., Gillberg, C., et al. (2003). Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism. Nat Genet.  Jamain, S., Radyushkin, K., Hammerschmidt, K., Granon, S., Boretius, S., Varoqueaux, F., Ramanantsoa, N., Gallego, J., Ronnenberg, A., Winter, D., et al. (2008). Reduced social interaction and ultrasonic communication in a mouse model of monogenic heritable autism. Proceedings of the National Academy of Sciences 105, 1710-1715.  Jontes, J. D., Emond, M. R., and Smith, S. J. ( 2004). In vivo trafficking and targeting of N-cadherin to nascent presynaptic terminals. J Neurosci 24, 9027-9034.  Kang, Y., Zhang, X., Dobie, F., Wu, H., and Craig, A. M. (2008). Induction of GABAergic Postsynaptic Differentiation by {alpha}-Neurexins. J Biol Chem 283, 2323-2334.  Kawakatsu, T., Shimizu, K., Honda, T., Fukuhara, T., Hoshino, T., and Takai, Y. (2002). Trans-interactions of nectins induce formation of filopodia and Lamellipodia through the respective activation of Cdc42 and Rac small G proteins. J Biol Chem 277, 50749-50755.  Kayser, M. S., McClelland, A. C., Hughes, E. G., and Dalva, M. B. (2006). Intracellular and Trans- Synaptic Regulation of Glutamatergic Synaptogenesis by EphB Receptors. J Neurosci 26, 12152-12164.  Keller-Peck, C. R., Walsh, M. K., Gan, W. B., Feng, G., Sanes, J. R., and Lichtman, J. W. (2001). Asynchronous synapse elimination in neonatal motor units: studies using GFP transgenic mice. Neuron 31, 381-394.  Kennedy, M. B. (1997). The postsynaptic density at glutamatergic synapses. Trends Neurosci 20, 264-268.   57 Khosravani, H., Altier, C., Zamponi, G. W., and Colicos, M. A. (2005). The Arg473Cys-neuroligin-1 mutation modulates NMDA mediated synaptic transmission and receptor distribution in hippocampal neurons. FEBS Letters 579, 6587-6594.  Kim, C. H., Chung, H. J., Lee, H. K., and Huganir, R. L. (2001). Interaction of the AMPA receptor subunit GluR2/3 with PDZ domains regulates hippocampal long-term depression. Proc Natl Acad Sci U S A 98, 11725-11730.  Kim, E., and Sheng, M. (2004). PDZ domain proteins of synapses. Nat Rev Neurosci 5, 771-781. Kim, H. G., Kishikawa, S., Higgins, A. W., Seong, I. S., Donovan, D. J., Shen, Y., Lally, E., Weiss, L. A.,  Najm, J., Kutsche, K., et al. (2008). Disruption of neurexin 1 associated with autism spectrum disorder. Am J Hum Genet 82, 199-207.  Kim, S., Burette, A., Chung, H. S., Kwon, S. K., Woo, J., Lee, H. W., Kim, K., Kim, H., Weinberg, R. J., and Kim, E. (2006). NGL family PSD-95-interacting adhesion molecules regulate excitatory synapse formation. Nat Neurosci 9, 1294-1301.  Kirsch, J., Langosch, D., Prior, P., Littauer, U. Z., Schmitt, B., and Betz, H. (1991). The 93-kDa glycine receptor-associated protein binds to tubulin. J Biol Chem 266, 22242-22245.  Kitano, J., Yamazaki, Y., Kimura, K., Masukado, T., Nakajima, Y., and Nakanishi, S. (2003). Tamalin is a scaffold protein that interacts with multiple neuronal proteins in distinct modes of protein-protein association. J Biol Chem 278, 14762-14768.  Kittler, J. T., and Moss, S. J. (2001). Neurotransmitter receptor trafficking and the regulation of synaptic strength. Traffic 2, 437-448.  Kneussel, M., and Betz, H. (2000a). Clustering of inhibitory neurotransmitter receptors at developing postsynaptic sites: the membrane activation model. Trends in Neurosciences 23, 429-435.  Kneussel, M., and Betz, H. (2000b). Clustering of inhibitory neurotransmitter receptors at developing postsynaptic sites: the membrane activation model. Trends Neurosci 23, 429-435.  Kneussel, M., Brandstatter, J. H., Gasnier, B., Feng, G., Sanes, J. R., and Betz, H. (2001). Gephyrin- independent clustering of postsynaptic GABA(A) receptor subtypes. Mol Cell Neurosci 17, 973-982.  Kneussel, M., Brandstatter, J. H., Laube, B., Stahl, S., Muller, U., and Betz, H. (1999). Loss of postsynaptic GABA(A) receptor clustering in gephyrin-deficient mice. J Neurosci 19, 9289-9297.  Kneussel, M., and Loebrich, S. (2007 ). Trafficking and synaptic anchoring of ionotropic inhibitory neurotransmitter receptors. Biology of the Cell 099, 297-309.  Knott, G. W., Quairiaux, C., Genoud, C., and Welker, E. (2002). Formation of dendritic spines with GABAergic synapses induced by whisker stimulation in adult mice. Neuron  Apr 11; 34, 265-273.  Knuesel, I., Bornhauser, B. C., Zuellig, R. A., Heller, F., Schaub, M. C., and Fritschy, J. M. (2000). Differential expression of utrophin and dystrophin in CNS neurons: an in situ hybridization and immunohistochemical study. J Comp Neurol 422, 594-611.  Knuesel, I., Mastrocola, M., Zuellig, R. A., Bornhauser, B., Schaub, M. C., and Fritschy, J. M. (1999). Short communication: altered synaptic clustering of GABAA receptors in mice lacking dystrophin (mdx mice). Eur J Neurosci 11, 4457-4462.   58 Koehnke, J., Jin, X., Budreck, E. C., Posy, S., Scheiffele, P., Honig, B., and Shapiro, L. (2008). Crystal structure of the extracellular cholinesterase-like domain from neuroligin-2. Proceedings of the National Academy of Sciences 105, 1873-1878.  Kolch, W. (2003). Erbin: sorting out ErbB2 receptors or giving Ras a break? Sci STKE 2003, pe37.  Konstantareas, M. M., and Homatidis, S. (1999). Chromosomal abnormalities in a series of children with autistic disorder. J Autism Dev Disord 29, 275-285.  Kornau, H. C., Schenker, L., T, , Kennedy, M. B., and Seeburg, P. H. (1995). Domain interaction between NMDA receptor subunits and the postsynaptic density protein PSD-95. Science 296, 1737-1740.  Kornau, H. C., Seeburg, P. H., and Kennedy, M. B. (1997). Interaction of ion channels and receptors with PDZ domain proteins. Curr Opin Neurobiol 7, 368-373.  Kwon, J. Y., Hong, M., Choi, M. S., Kang, S., Duke, K., Kim, S., Lee, S., and Lee, J. (2004). Ethanol- response genes and their regulation analyzed by a microarray and comparative genomic approach in the nematode Caenorhabditis elegans. Genomics 83, 600-614.  Lahey, T., Gorczyca, M., Jia, X. X., and Budnik, V. (1994). The Drosophila tumor suppressor gene dlg is required for normal synaptic bouton structure. Neuron 13, 823-835.  Lardi-Studler, B., and Fritschy, J.-M. (2007). Matching of Pre- and Postsynaptic Specializations during Synaptogenesis. Neuroscientist 13, 115-126.  Laumonnier, F., Bonnet-Brilhault, F., Gomot, M., Blanc, R., David, A., Moizard, M. P., Raynaud, M., Ronce, N., Lemonnier, E., Calvas, P., et al. (2004). X-Linked Mental Retardation and Autism Are Associated with a Mutation in the NLGN4 Gene, a Member of the Neuroligin Family. Am J Hum Genet 74, 552-557.  Laurie, D., Wisden, W., and Seeburg, P. (1992). The distribution of thirteen GABAA receptor subunit mRNAs in the rat brain. III. Embryonic and postnatal development. J Neurosci 12, 4151-4172.  Lawson-Yuen, A., Saldivar, J.-S., Sommer, S., and Picker, J. (2008). Familial deletion within NLGN4 associated with autism and Tourette syndrome.  Lee, R. C., Clandinin, T. R., Lee, C.-H., Chen, P.-L., Meinertzhagen, I. A., and Zipursky, S. L. (2003). The protocadherin Flamingo is required for axon target selection in the Drosophila visual system. 6, 557-563.  Levi, S., Grady, R. M., Henry, M. D., Campbell, K. P., Sanes, J. R., and Craig, A. M. (2002). Dystroglycan is selectively associated with inhibitory GABAergic synapses but is dispensable for their differentiation. J Neurosci 22, 4274-4285.  Levi, S., Logan, S. M., Tovar, K. R., and Craig, A. M. (2004). Gephyrin is critical for glycine receptor clustering but not for the formation of functional GABAergic synapses in hippocampal neurons. J Neurosci 24, 207-217.  Levinson, J. N., Chery, N., Huang, K., Wong, T. P., Gerrow, K., Kang, R., Prange, O., Wang, Y. T., and El-Husseini, A. (2005). Neuroligins Mediate Excitatory and Inhibitory Synapse Formation: INVOLVEMENT OF PSD-95 AND NEUREXIN-1{beta} IN NEUROLIGIN-INDUCED SYNAPTIC SPECIFICITY. J Biol Chem 280, 17312-17319.  Levinson, J. N., and El-Husseini, A. (2005a). Building Excitatory and Inhibitory Synapses: Balancing Neuroligin Partnerships. Neuron 48, 171-174.   59 Levinson, J. N., and El-Husseini, A. (2005b). New players tip the scales in the balance between excitatory and inhibitory synapses. Mol Pain 1, 12.  Li, J., Ashley, J., Budnik, V., and Bhat, M. A. (2007). Crucial role of Drosophila neurexin in proper active zone apposition to postsynaptic densities, synaptic growth, and synaptic transmission. Neuron 55, 741-755.  Li, R. W., Serwanski, D. R., Miralles, C. P., Li, X., Charych, E., Riquelme, R., Huganir, R. L., and de Blas, A. L. (2005). GRIP1 in GABAergic synapses. J Comp Neurol 488, 11-27.  Lichtman, J. W., and Colman, H. (2000). Synapse elimination and indelible memory. Neuron 25, 269-278.  Lin, D.-T., and Huganir, R. L. (2007). PICK1 and Phosphorylation of the Glutamate Receptor 2 (GluR2) AMPA Receptor Subunit Regulates GluR2 Recycling after NMDA Receptor-Induced Internalization. J Neurosci 27, 13903-13908.  Lin, W., Burgess, R.W., Dominguez, B., Pfaff, S.L., Sanes, J.R., Lee, K.F. (2001). Distinct roles of nerve and muscle in postsynaptic differentiation of the neuromuscular synapse. Nature 410, 1057-1064.  Lisé, M. F., and El-Husseini, A. (2006). The neuroligin and neurexin families: from structure to function at the synapse. Cell Mol Life Sci 63, 1833-1849.  Malinow, R., and Malenka, R. C. (2002). AMPA receptor trafficking and synaptic plasticity. Annu Rev Neurosci 25, 103-126.  Mariner, R., Jackson, A. W. r., Levitas, A., Hagerman, R. J., Braden, M., McBogg, P. M., Smith, A. C., and Berry, R. (1986). Autism, mental retardation, and chromosomal abnormalities. J Autism Dev Disord 16, 425-440.  Marrs, G. S., Green, S. H., and Dailey, M. E. (2001). Rapid formation and remodeling of postsynaptic densities in developing dendrites. Nat Neurosci 4, 1006-1013.  Marshall, C. R., Noor, A., Vincent, J. B., Lionel, A. C., Feuk, L., Skaug, J., Shago, M., Moessner, R., Pinto, D., Ren, Y., et al. (2008). Structural Variation of Chromosomes in Autism Spectrum Disorder. The American Journal of Human Genetics 82, 477-488.  Matteoli, M., Takei, K., Perin, M., Sudhof, T., and De Camilli, P. (1992). Exo-endocytotic recycling of synaptic vesicles in developing processes of cultured hippocampal neurons. J Cell Biol 117, 849-861.  McAllister, A. K. (2007). Dynamic Aspects of CNS Synapse Formation. Annual Review of Neuroscience 30, 425-450.  McMahan, U. J. (1990). The agrin hypothesis. Cold Spring Harb Symp Quant Biol 55, 407-418.  Meier, J., Vannier, C., Serge, A., Triller, A., and Choquet, D. (2001). Fast and reversible trapping of surface glycine receptors by gephyrin. Nat Neurosci 4, 253-260.  Missler, M., Fernandez-Chacon, R., and Sudhof, T. C. (1998). The making of neurexins. J Neurochem 71, 1339-1347.  Missler, M., and Sudhof, T. C. (1998). Neurexins: three genes and 1001 products. Trends in Genetics 14, 20-26.  Mizoguchi, A., Nakanishi, H., Kimura, K., Matsubara, K., Ozaki-Kuroda, K., Katata, T., Honda, T., Kiyohara, Y., Heo, K., Higashi, M., et al. (2002). Nectin: an adhesion molecule involved in formation of synapses. J Cell Biol 156, 555-565.   60 Moessner, R., Marshall, C. R., Sutcliffe, J. S., Skaug, J., Pinto, D., Vincent, J., Zwaigenbaum, L., Fernandez, B., Roberts, W., Szatmari, P., and Scherer, S. W. (2007). Contribution of SHANK3 mutations to autism spectrum disorder. Am J Hum Genet 81, 1289-1297.  Montgomery, J. M., Zamorano, P. L., and Garner, C. C. (2004). MAGUKs in synapse assembly and function: an emerging view. Cell Mol Life Sci 61, 911-929.  Moss, S. J., and Smart, T. G. (2001). Constructing inhibitory synapses. Nat Rev Neurosci 2, 240-250. Muller, B. M., Kistner, U., Kindler, S., Chung, W. J., Kuhlendahl, S., Fenster, S. D., Lau, L. F., Veh, R. W., Huganir, R. L., Gundelfinger, E. D., and Garner, C. C. (1996). SAP102, a novel postsynaptic protein that interacts with NMDA receptor complexes in vivo. Neuron 17, 255-265.  Nam, C. I., and Chen, L. (2005). Postsynaptic assembly induced by neurexin-neuroligin interaction and neurotransmitter. Proc Natl Acad Sci U S A 102, 6137-6142.  Nguyen, T., and Sudhof, T. C. (1997). Binding Properties of Neuroligin 1 and Neurexin 1beta  Reveal Function as Heterophilic Cell Adhesion Molecules. J Biol Chem 272, 26032-26039.  Nimchinsky, E. A., Sabatini, B. L., and Svoboda, K. (2002). Structure and function of dendritic spines. Annu Rev Physiol 64, 313-353.  Nishimura, W., Yao, I., Iida, J., Tanaka, N., and Hata, Y. (2002). Interaction of synaptic scaffolding molecule and Beta -catenin. J Neurosci 22, 757-765.  Noakes, P. G., Gautam, M., Mudd, J., Sanes, J. R., and Merlie, J. P. (1995). Aberrant differentiation of neuromuscular junctions in mice lacking s-laminin/laminin beta 2. Nature 374, 258-262.  Nusser, Z., Sieghart, W., Benke, D., Fritschy, J.-M., and Somogyi, P. (1996). Differential synaptic localization of two major gamma -aminobutyric acid type A receptor alpha  subunits on hippocampal pyramidal cells. Proceedings of the National Academy of Sciences 93, 11939-11944.  O'Brien, R. J., Lau, L. F., and Huganir, R. L. (1998). Molecular mechanisms of glutamate receptor clustering at excitatory synapses. Curr Opin Neurobiol 8, 364-369.  O'Brien, R. J., Xu, D., Petralia, R. S., Steward, O., Huganir, R. L., and Worley, P. (1999). Synaptic clustering of AMPA receptors by the extracellular immediate-early gene product Narp. Neuron 23, 195- 197.  Ohtsuka, T., Takao-Rikitsu, E., Inoue, E., Inoue, M., Takeuchi, M., Matsubara, K., Deguchi-Tawarada, M., Satoh, K., Morimoto, K., Nakanishi, H., and Takai, Y. (2002). Cast: a novel protein of the cytomatrix at the active zone of synapses that forms a ternary complex with RIM1 and munc13-1. J Cell Biol 158, 577-590.  Okabe, S. (2007). Molecular anatomy of the postsynaptic density. Molecular and Cellular Neuroscience 34, 503-518.  Okabe, S., Miwa, A., and Okado, H. (2001). Spine formation and correlated assembly of presynaptic and postsynaptic molecules. J Neurosci 21, 6105-6114.  O'Sullivan, G. A., Kneussel, M., Elazar, Z., and Betz, H. (2005). GABARAP is not essential for GABA receptor targeting to the synapse. Eur J Neurosci 22, 2644-2648.  Panzer, J. A., Song, Y., and Balice-Gordon, R. J. (2006). In Vivo Imaging of Preferential Motor Axon Outgrowth to and Synaptogenesis at Prepatterned Acetylcholine Receptor Clusters in Embryonic Zebrafish Skeletal Muscle. J Neurosci 26, 934-947.   61 Paraoanu, L. E., Becker-Roeck, M., Christ, E., and Layer, P. G. (2006). Expression patterns of neurexin-1 and neuroligins in brain and retina of the chick embryo: Neuroligin-3 is absent in retina. Neuroscience Letters 395, 114-117.  Passafaro, M., Nakagawa, T., Sala, C., and Sheng, M. (2003). Induction of dendritic spines by an extracellular domain of AMPA receptor subunit GluR2. Nature 424, 677-681.  Passafaro, M., Piëch, V., and Sheng, M. (2001). Subunit-specific temporal and spatial patterns of AMPA receptor exocytosis in hippocampal neurons. 4, 917-926.  Patterson, S. I. (2002). Posttranslational protein S-palmitoylation and the compartmentalization of signaling molecules in neurons. Biol Res 35, 139-150.  Patton, B. L., Cunningham, J. M., Thyboll, J., Kortesmaa, J., Westerblad, H., Edstrom, L., Tryggvason, K., and Sanes, J. R. (2001). Properly formed but improperly localized synaptic specializations in the absence of laminin alpha4. Nat Neurosci 4, 597-604.  Peng, J., Kim, M. J., Cheng, D., Duong, D. M., Gygi, S. P., and Sheng, M. (2004). Semiquantitative proteomic analysis of rat forebrain postsynaptic density fractions by mass spectrometry. J Biol Chem 279, 21003-21011.  Perroy, J., El Far, O., Bertaso, F., Pin, J. P., Betz, H., Bockaert, J., and Fagni, L. (2002). PICK1 is required for the control of synaptic transmission by the metabotropic glutamate receptor 7. Embo J 21, 2990-2999.  Petralia, R. S., Sans, N., Wang, Y.-X., and Wenthold, R. J. (2005). Ontogeny of postsynaptic density proteins at glutamatergic synapses. Molecular and Cellular Neuroscience 29, 436-452.  Pierce, J. P., and Mendell, L. M. (1993). Quantitative ultrastructure of Ia boutons in the ventral horn: scaling and positional relationships. J Neurosci 13, 4748-4763.  Prange, O., and Murphy, T. H. (2001). Modular transport of postsynaptic density-95 clusters and association with stable spine precursors during early development of cortical neurons. J Neurosci 21, 9325- 9333.  Prange, O., Wong, T. P., Gerrow, K., Wang, Y. T., and El-Husseini, A. (2004). A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. PNAS 101, 13915-13920.  Puschel, A., and Betz, H. (1995). Neurexins are differentially expressed in the embryonic nervous system of mice. J Neurosci 15, 2849-2856.  Rao, A., Cha, E. M., and Craig, A. M. (2000). Mismatched appositions of presynaptic and postsynaptic components in isolated hippocampal neurons. J Neurosci 20, 8344-8353.  Rao, A., and Craig, A. M. (1997). Activity regulates the synaptic localization of the NMDA receptor in hippocampal neurons. Neuron 19, 801-812.  Rao, A., Kim, E., Sheng, M., and Craig, A. M. (1998). Heterogeneity in the molecular composition of excitatory postsynaptic sites during development of hippocampal neurons in culture. J Neurosci 18, 1217- 1229.  Ribchester, R. R., and Taxt, T. (1983). Motor unit size and synaptic competition in rat lumbrical muscles reinnervated by active and inactive motor axons. J Physiol 344, 89-111.  Risch, N., Spiker, D., Lotspeich, L., Nouri, N., Hinds, D., Hallmayer, J., Kalaydjieva, L., McCague, P., Dimiceli, S., Pitts, T., et al. (1999). A genomic screen of autism: evidence for a multilocus etiology. Am J Hum Genet 65, 493-507.  62  Rosales, C. R., Osborne, K. D., Zuccarino, G. V., Scheiffele, P., and Silverman, M. A. (2005). A cytoplasmic motif targets neuroligin-1 exclusively to dendrites of cultured hippocampal neurons. European Journal of Neuroscience 22, 2381-2386.  Rosenmund, C., Rettig, J., and Brose, N. (2003). Molecular mechanisms of active zone function. Current Opinion in Neurobiology 13, 509-519.  Sabo, S. L., Gomes, R. A., and McAllister, A. K. (2006). Formation of Presynaptic Terminals at Predefined Sites along Axons. J Neurosci 26, 10813-10825.  Sabo, S. L., and McAllister, A. K. (2003). Mobility and cycling of synaptic protein-containing vesicles in axonal growth cone filopodia. Nat Neurosci 6, 1264-1269.  Saghatelyan, A. K., Nikonenko, A. G., Sun, M., Rolf, B., Putthoff, P., Kutsche, M., Bartsch, U., Dityatev, A., and Schachner, M. (2004). Reduced GABAergic transmission and number of hippocampal perisomatic inhibitory synapses in juvenile mice deficient in the neural cell adhesion molecule L1. Mol Cell Neurosci 26, 191-203.  Saglietti, L., Dequidt, C., Kamieniarz, K., Rousset, M. C., Valnegri, P., Thoumine, O., Beretta, F., Fagni, L., Choquet, D., Sala, C., et al. (2007). Extracellular interactions between GluR2 and N-cadherin in spine regulation. Neuron 54, 461-477.  Sala, C., Piech, V., Wilson, N. R., Passafaro, M., Liu, G., and Sheng, M. (2001). Regulation of dendritic spine morphology and synaptic function by Shank and Homer. Neuron 31, 115-130.  Sanes, J. R., and Lichtman, J. W. (1999). Development of the vertebrate neuromuscular junction. Annu Rev Neurosci 22, 389-442.  Sans, N., Petralia, R. S., Wang, Y. X., Blahos, J., 2nd, Hell, J. W., and Wenthold, R. J. (2000). A developmental change in NMDA receptor-associated proteins at hippocampal synapses. J Neurosci 20, 1260-1271.  Scheiffele, P., Fan, J., Choih, J., Fetter, R., and Serafini, T. (2000). Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell 101, 657-669.  Schneggenburger, R., and Neher, E. (2005). Presynaptic calcium and control of vesicle fusion. Current Opinion in Neurobiology Signalling mechanisms 15, 266-274.  Scholl, F. G., and Scheiffele, P. (2003). Making connections: cholinesterase-domain proteins in the CNS. Trends in Neurosciences 26, 618-624.  Shapira, M., Zhai, R. G., Dresbach, T., Bresler, T., Torres, V. I., Gundelfinger, E. D., Ziv, N. E., and Garner, C. C. (2003). Unitary assembly of presynaptic active zones from Piccolo-Bassoon transport vesicles. Neuron 38, 237-252.  Sheckler, L. R., Henry, L., Sugita, S., Sudhof, T. C., and Rudenko, G. (2006). Crystal Structure of the Second LNS/LG Domain from Neurexin 1{alpha}: Ca2+ BINDING AND THE EFFECTS OF ALTERNATIVE SPLICING. J Biol Chem 281, 22896-22905.  Shen, K., Fetter, R. D., and Bargmann, C. I. (2004). Synaptic specificity is generated by the synaptic guidepost protein SYG-2 and its receptor, SYG-1. Cell 116, 869-881.  Sheng, M. (1996). PDZs and receptor/channel clustering: rounding up the latest suspects. Neuron 17, 575- 578.   63 Sheng, M., and Sala, C. (2001). PDZ domains and the organization of supramolecular complexes. Annu Rev Neurosci 24, 1-29.  Sia, G. M., Béïque, J. C., Rumbaugh, G., Cho, R., Worley, P. F., and Huganir, R. L. (2007). Interaction of the N-Terminal Domain of the AMPA Receptor GluR4 Subunit with the Neuronal Pentraxin NP1 Mediates GluR4 Synaptic Recruitment. Neuron 55, 87-102.  Smith, M. A., Gergans, S. R., and Lyle, M. A. (2006). The motor-impairing effects of GABA(A) and GABA(B) agonists in gamma-hydroxybutyrate (GHB)-treated rats: cross-tolerance to baclofen but not flunitrazepam. Eur J Pharmacol 552, 83-89.  Sola, M., Bavro, V. N., Timmins, J., Franz, T., Ricard-Blum, S., Schoehn, G., Ruigrok, R. W., Paarmann, I., Saiyed, T., O'Sullivan, G. A., et al. (2004). Structural basis of dynamic glycine receptor clustering by gephyrin. Embo J 23, 2510-2519.  Sola, M., Kneussel, M., Heck, I. S., Betz, H., and Weissenhorn, W. (2001). X-ray crystal structure of the trimeric N-terminal domain of gephyrin. J Biol Chem 276, 25294-25301.  Song, J. Y., Ichtchenko, K., Sudhof, T. C., and Brose, N. (1999). Neuroligin 1 is a postsynaptic cell- adhesion molecule of excitatory synapses. Proc Natl Acad Sci U S A 96, 1100-1105.  Sons, M. S., Busche, N., Strenzke, N., Moser, T., Ernsberger, U., Mooren, F. C., Zhang, W., Ahmad, M., Steffens, H., Schomburg, E. D., et al. (2006). [alpha]-Neurexins are required for efficient transmitter release and synaptic homeostasis at the mouse neuromuscular junction. Neuroscience 138, 433-446.  Sperry, R. W. (1963). Chemoaffinity in the orderly growth of nerve fibre patterns and connections. Proc Natl Acad Sci U S A 50, 703-710.  Sretavan, D. W., and Shatz, C. J. (1986). Prenatal development of cat retinogeniculate axon arbors in the absence of binocular interactions. J Neurosci 6, 990-1003.  Srivastava, S., Osten, P., Vilim, F. S., Khatri, L., Inman, G., States, B., Daly, C., DeSouza, S., Abagyan, R., Valtschanoff, J. G., et al. (1998). Novel anchorage of GluR2/3 to the postsynaptic density by the AMPA receptor-binding protein ABP. Neuron 21, 581-591.  Srivastava, S., and Ziff, E. B. (1999). ABP: a novel AMPA receptor binding protein. Ann N Y Acad Sci 868, 561-564.  Stein, V., House, D. R., Bredt, D. S., and Nicoll, R. A. (2003). Postsynaptic density-95 mimics and occludes hippocampal long-term potentiation and enhances long-term depression. J Neurosci 23, 5503- 5506.  Steinberg, J. P., Takamiya, K., Shen, Y., Xia, J., Rubio, M. E., Yu, S., Jin, W., Thomas, G. M., Linden, D. J., and Huganir, R. L. (2006). Targeted In Vivo Mutations of the AMPA Receptor Subunit GluR2 and Its Interacting Protein PICK1 Eliminate Cerebellar Long-Term Depression. Neuron 49, 845-860.  Stevens, C. F. (2003). Neurotransmitter Release at Central Synapses. Neuron 40, 381-388. Studler, B., Fritschy, J. M., and Brunig, I. (2002). GABAergic and glutamatergic terminals differentially influence the organization of GABAergic synapses in rat cerebellar granule cells in vitro. Neuroscience 114, 123-133.  Sugita, S., Saito, F., Tang, J., Satz, J., Campbell, K., and Sudhof, T. C. (2001). A stoichiometric complex of neurexins and dystroglycan in brain. J Cell Biol 154, 435-445.  Sugiyama, J., Bowen, D. C., and Hall, Z. W. (1994). Dystroglycan binds nerve and muscle agrin. Neuron 13, 103-115.  64  Sumita, K., Sato, Y., Iida, J., Kawata, A., Hamano, M., Hirabayashi, S., Ohno, K., Peles, E., and Hata, Y. (2007). Synaptic scaffolding molecule (S-SCAM) membrane-associated guanylate kinase with inverted organization (MAGI)-2 is associated with cell adhesion molecules at inhibitory synapses in rat hippocampal neurons. Journal of Neurochemistry 100, 154-166.  Sunderland, W. J., Son, Y. J., Miner, J. H., Sanes, J. R., and Carlson, S. S. (2000). The presynaptic calcium channel is part of a transmembrane complex linking a synaptic laminin (alpha4beta2gamma1) with non- erythroid spectrin. J Neurosci 20, 1009-1019.  Takahashi, H., Sekino, Y., Tanaka, S., Mizui, T., Kishi, S., and Shirao, T. (2003). Drebrin-Dependent Actin Clustering in Dendritic Filopodia Governs Synaptic Targeting of Postsynaptic Density-95 and Dendritic Spine Morphogenesis. J Neurosci 23, 6586-6595.  Takai, Y., and Nakanishi, H. (2003). Nectin and afadin: novel organizers of intercellular junctions. J Cell Sci 116, 17-27.  Taniguchi, H., Gollan, L., Scholl, F. G., Mahadomrongkul, V., Dobler, E., Limthong, N., Peck, M., Aoki, C., and Scheiffele, P. (2007). Silencing of Neuroligin Function by Postsynaptic Neurexins. J Neurosci 27, 2815-2824.  Tardin, C., Cognet, L., Bats, C., Lounis, B., and Choquet, D. (2003). Direct imaging of lateral movements of AMPA receptors inside synapses. Embo J 22, 4656-4665.  Thomas, N. S., Sharp, A. J., Browne, C. E., Skuse, D., Hardie, C., and Dennis, N. R. (1999). Xp deletions associated with autism in three females. Human Genetics 104, 43-48.  Thompson, W. J. (1985). Activity and synapse elimination at the neuromuscular junction. Cell Mol Neurobiol 5, 167-182.  Togashi, H., Abe, K., Mizoguchi, A., Takaoka, K., Chisaka, O., and Takeichi, M. (2002). Cadherin regulates dendritic spine morphogenesis. Neuron 35, 77-89.  Trachtenberg, J. T., B.E., C., G.W., K., Feng, G., Sanes, J. R., Welker, E., and Svoboda, K. (2002). Long- term in vivo imaging of experience-dependent synaptic plasticity in adult cortex. Nature 420, 788-794.  Tretter, V., Jacob, T. C., Mukherjee, J., Fritschy, J.-M., Pangalos, M. N., and Moss, S. J. (2008). The Clustering of GABAA Receptor Subtypes at Inhibitory Synapses is Facilitated via the Direct Binding of Receptor {alpha}2 Subunits to Gephyrin. J Neurosci 28, 1356-1365.  Tsigelny, I., Shindyalov, I., Bourne, P., Sudhof, T., and Taylor, P. (2000). Common EF-hand motifs in cholinesterases and neuroligins suggest a role for Ca2+ binding in cell surface associations. Protein Sci 9, 180-185.  Tu, J. C., Xiao, B., Naisbitt, S., Yuan, J. P., Petralia, R. S., Brakeman, P., Doan, A., Aakalu, V. K., Lanahan, A. A., Sheng, M., and Worley, P. F. (1999). Coupling of mGluR/Homer and PSD-95 complexes by the Shank family of postsynaptic density proteins. Neuron 23, 583-592.  Tyc, F., and Vrbova, G. (1995). Stabilisation of neuromuscular junctions by leupeptin increases motor unit size in partially denervated rat muscles. Brain Res Dev Brain Res 88, 186-193.  Ullrich, B., Ushkaryov, Y. A., and Sudhof, T. C. (1995). Cartography of neurexins: more than 1000 isoforms generated by alternative splicing and expressed in distinct subsets of neurons. Neuron 14, 497- 507.   65 Ushkaryov, Y. A., Hata, Y., Ichtchenko, K., Moomaw, C., Afendis, S., Slaughter, C. A., and Sudhof, T. C. (1994). Conserved domain structure of beta-neurexins. Unusual cleaved signal sequences in receptor-like neuronal cell-surface proteins. J Biol Chem 269, 11987-11992.  VanSaun, M., Herrera, A. A., and Werle, M. J. (2003). Structural alterations at the neuromuscular junctions of matrix metalloproteinase 3 null mutant mice. J Neurocytol 32, 1129-1142.  Varoqueaux, F., Jamain, S., and Brose, N. (2004). Neuroligin 2 is exclusively localized to inhibitory synapses. Eur J Cell Biol 83, 449-456.  Vaughn, J. E. (1989). Fine structure of synaptogenesis in the vertebrate central nervous system. Synapse 3, 255-285.  Vigot, R., Barbieri, S., Bräuner-Osborne, H., Turecek, R., Shigemoto, R., Zhang, Y. P., Luján, R., Jacobson, L. H., Biermann, B., Fritschy, J. M., et al. (2006). Differential compartmentalization and distinct functions of GABAB receptor variants. Neuon 50, 589-601.  Vincent, J. B., Kolozsvari, D., Roberts, W. S., Bolton, P. F., Gurling, H. M. D., and Scherer, S. W. (2004). Mutation screening of X-chromosomal neuroligin genes: No mutations in 196 autism probands. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 129B, 82-84.  Vock, V. M., Ponomareva, O. N., and Rimer, M. (2008). Evidence for Muscle-Dependent Neuromuscular Synaptic Site Determination in Mammals. J Neurosci 28, 3123-3130.  Vogt, A. K., Brewer, G. J., Decker, T., Bocker-Meffert, S., Jacobsen, V., Kreiter, M., Knoll, W., and Offenhausser, A. (2005). Independence of synaptic specificity from neuritic guidance. Neuroscience 134, 783-790.  Waites, C. L., Craig, A. M., and Garner, C. C. (2005a). Mechanisms of vertebrate synaptogenesis. Annu Rev Neurosci 28, 251-274.  Waites, C. L., Craig, A. M., and Garner, C. C. (2005b). MECHANISMS OF VERTEBRATE SYNAPTOGENESIS. Annual Review of Neuroscience 28, 251-274.  Walikonis, R. S., Jensen, O. N., Mann, M., Provance, D. W., Jr., Mercer, J. A., and Kennedy, M. B. (2000). Identification of proteins in the postsynaptic density fraction by mass spectrometry. J Neurosci 20, 4069- 4080.  Walikonis, R. S., Oguni, A., Khorosheva, E. M., Jeng, C. J., Asuncion, F. J., and Kennedy, M. B. (2001). Densin-180 forms a ternary complex with the (alpha)-subunit of Ca2+/calmodulin-dependent protein kinase II and (alpha)-actinin. J Neurosci 21, 423-433.  Wang, C.-Y., Chang, K., Petralia, R. S., Wang, Y.-X., Seabold, G. K., and Wenthold, R. J. (2006). A Novel Family of Adhesion-Like Molecules That Interacts with the NMDA Receptor. J Neurosci 26, 2174-2183.  Wang, H., Bedford, F. K., Brandon, N. J., Moss, S. J., and Olsen, R. W. (1999). GABA(A)-receptor- associated protein links GABA(A) receptors and the cytoskeleton. Nature 397, 69-72.  Wang, H., and Olsen, R. W. (2000). Binding of the GABA(A) receptor-associated protein (GABARAP) to microtubules and microfilaments suggests involvement of the cytoskeleton in GABARAPGABA(A) receptor interaction. J Neurochem 75, 644-655.  Washbourne, P., Bennett, J. E., and McAllister, A. K. (2002). Rapid recruitment of NMDA receptor transport packets to nascent synapses. Nat Neurosci 5, 751-759.   66 Washbourne, P., Liu, X. B., Jones, E. G., and McAllister, A. K. (2004). Cycling of NMDA receptors during trafficking in neurons before synapse formation. J Neurosci 24, 8253-8264.  Wassink, T. H., Brzustowicz, L. M., Bartlett, C. W., and Szatmari, P. (2004). The search for autism disease genes. Mental Retardation and Developmental Disabilities Research Reviews 10, 272-283.  Weiner, J. A., Wang, X., Tapia, J. C., and Sanes, J. R. (2005). Gamma protocadherins are required for synaptic development in the spinal cord. Proc Natl Acad Sci U S A 102, 8-14.  Wenthold, R. J., Prybylowski, K., Standley, S., Sans, N., and Petralia, R. S. (2003). Trafficking of NMDA receptors. Annu Rev Pharmacol Toxicol 43, 335-358.  WHITING, P. J., BONNERT, T. P., MCKERNAN, R. M., FARRAR, S., BOURDELLES, B. L., HEAVENS, R. P., SMITH, D. W., HEWSON, L., RIGBY, M. R., SIRINATHSINGHJI, D. J. S., et al. (1999). Molecular and Functional Diversity of the Expanding GABA-A Receptor Gene Family. Ann NY Acad Sci 868, 645-653.  Williams, M. E., Wu, S. C., McKenna, W. L., and Hinck, L. (2003). Surface expression of the netrin receptor UNC5H1 is regulated through a protein kinase C-interacting protein/protein kinase-dependent mechanism. J Neurosci 23, 11279-11288.  Williamson, R. A., Henry, M. D., Daniels, K. J., Hrstka, R. F., Lee, J. C., Sunada, Y., Ibraghimov- Beskrovnaya, O., and Campbell, K. P. (1997). Dystroglycan is essential for early embryonic development: disruption of Reichert's membrane in Dag1-null mice. Hum Mol Genet 6, 831-841.  Wilson, D. A., Best, A. R., and Sullivan, R. M. (2004). Plasticity in the olfactory system: lessons for the neurobiology of memory. Neuroscientist 10, 513-524.  Wyszynski, M., Kim, E., Dunah, A. W., Passafaro, M., Valtschanoff, J. G., Serra-Pagès, C., Streuli, M., Weinberg, R. J., and Sheng, M. (2002). Interaction between GRIP and liprin-alpha/SYD2 is required for AMPA receptor targeting. Neuon 34, 39-52.  Xia, J., Zhang, X., Staudinger, J., and Huganir, R. L. (1999). Clustering of AMPA receptors by the synaptic PDZ domain-containing protein PICK1. Neuron 22, 179-187.  Yamagata, M., Weiner, J. A., and Sanes, J. R. (2002). Sidekicks: synaptic adhesion molecules that promote lamina-specific connectivity in the retina. Cell 110, 649-660.  Yan, J., Oliveira, G., Coutinho, A., Yang, C., Feng, J., Katz, C., Sram, J., Bockholt, A., Jones, I. R., Craddock, N., et al. (2004). Analysis of the neuroligin 3 and 4 genes in autism and other neuropsychiatric patients. Mol Psychiatry.  Yang, X., Arber, S., William, C., Li, L., Tanabe, Y., Jessell, T.M., Birchmeier, C., Burden, S.J. (2001). Patterning of muscle acetylcholine receptor gene expression in the absence of motor innervation. Neuon 30, 399-410.  Ye, B., Liao, D., Zhang, X., Zhang, P., Dong, H., and Huganir, R. L. (2000). GRASP-1: a neuronal RasGEF associated with the AMPA receptor/GRIP complex. Neuron 26, 603-617.  Ylisaukko-oja, T., Rehnstrom, K., Auranen, M., Vanhala, R., Alen, R., Kempas, E., Ellonen, P., Turunen, J. A., Makkonen, I., Riikonen, R., et al. (2005). Analysis of four neuroligin genes as candidates for autism.  Yu, W., Jiang, M., Miralles, C. P., Li, R.-w., Chen, G., and de Blas, A. L. (2007). Gephyrin clustering is required for the stability of GABAergic synapses. Molecular and Cellular Neuroscience 36, 484-500.   67 Zeng, X., Sun, M., Liu, L., Chen, F., Wei, L., and Xie, W. (2007). Neurexin-1 is required for synapse formation and larvae associative learning in Drosophila. FEBS letters 581, 2509-2516.  Zhai, R. G., Vardinon-Friedman, H., Cases-Langhoff, C., Becker, B., Gundelfinger, E. D., Ziv, N. E., and Garner, C. C. (2001). Assembling the presynaptic active zone: a characterization of an active one precursor vesicle. Neuron 29, 131-143.  Zhang, W., and Benson, D. L. (2000). Development and molecular organization of dendritic spines and their synapses. Hippocampus 10, 512-526.  Zhen, M., and Jin, Y. (2004). Presynaptic terminal differentiation: transport and assembly. Current Opinion in Neurobiology 14, 280-287.  Zhou, H., T Muramatsu, W Halfter, KW Tsim, HB Peng (1997). A role of midline in the development of the neuromuscular junction. Mol Cell Neurosci 10, 56-70.  Zhu, H., Wu, F., and Schacher, S. (1995). Changes in expression and distribution of Aplysia cell adhesion molecules can influence synapse formation and elimination in vitro. J Neurosci 1995 Jun;15(6):4173-83 15, 4173-4183.  Ziff, E. B. (1997). Enlightening the postsynaptic density. Neuron 19, 1163-1174.  Ziv, N. E. (2001). Recruitment of synaptic molecules during synaptogenesis. Neuroscientist 7, 365-370.  Ziv, N. E., and Garner, C. C. (2001). Principles of glutamatergic synapse formation: seeing the forest for the trees. Curr Opin Neurobiol 11, 536-543.  Ziv, N. E., and Garner, C. C. (2004). Cellular and molecular mechanisms of presynaptic assembly. Nat Rev Neurosci 5, 385-399.  Ziv, N. E., and Smith, S. J. (1996). Evidence for a role of dendritic filopodia in synaptogenesis and spine formation. Neuron 17, 91-102.    68 2 A preformed complex of postsynaptic scaffolding proteins is involved in glutamatergic synapse development                                    A version of this chapter has been published. Gerrow K, Romorini S, Nabi SM, Colicos MA, Sala C, El-Husseini A. (2006) A preformed complex of postsynaptic proteins is involved in excitatory synapse development. Neuron. 2006 Feb 16; 49(4):547-62.  69 2.1 INTRODUCTION  Communication between neuronal cells relies on signals transmitted through an array of heterogeneous contacts known as synapses (Sanes and Lichtman 2001; Li and Sheng 2003; Waites et al. 2005).  This process involves the establishment of thousands of excitatory and inhibitory synapses, each with characteristic components that determine synapse specificity. Excitatory neurotransmission in the mammalian brain is primarily mediated by glutamate and the proteinaceous network of neurotransmitter receptors, scaffolding proteins, adhesion molecules and signal transduction enzymes at glutamatergic synapses is collectively referred to as the postsynaptic density (PSD) (Kennedy 1997). Scaffolding proteins constitute one major group of proteins present at the PSD. Many of these molecules contain multiple domains for protein-protein interaction, such as PDZ domains and thus mediate the recruitment of several other proteins to a single cellular site (Rao et al. 1998; Kim and Sheng 2004).  The postsynaptic density protein PSD-95 is believed to play a role in synapse maturation, as it is one of the earliest detectable proteins in the PSD (Rao et al. 1998; Kim and Sheng 2004). PSD-95 induces clustering of a number of neurotransmitter receptors and scaffolding proteins (El-Husseini et al. 2000; Kim and Sheng 2004; Prange et al. 2004). PSD-95 mutant animals exhibit defects in synaptic transmission associated with plasticity and results in enhanced LTP and impaired learning (Migaud et al. 1998).  Moreover, knock down of PSD-95 diminishes excitatory synapse number and clustering of AMPA- type glutamate receptors (Nakagawa et al. 2004; Prange et al. 2004). Interaction of the scaffolding protein, GKAP, with PSD-95 is thought to be important for coupling of GKAP to Shank, a protein implicated in regulation of spine morphology (Sala et al. 2001). PSD-95 may control the balance of excitatory and inhibitory synaptic contacts through clustering of the cell adhesion molecule neuroligin-1 (Prange et al. 2004; Levinson et al. 2005). These findings suggest that PSD-95 has the potential to drive the formation of a core postsynaptic protein complex containing essential scaffolding and cell adhesion proteins needed to coordinate the formation and maturation of pre- and postsynaptic elements, as well as dictate their number and specificity.  70  Non-synaptic clusters of postsynaptic proteins in young neurons have been documented in several studies (Rao et al, 1998; Liu et al., 1998; Washbourne et al., 2002 ; Washbourneet al., 2004; Sans et al., 2003; Waites et al., 2005). Clusters containing PSD- 95 and GKAP present at non-synaptic sites have been observed in hippocampal neurons in vitro (Rao et al. 1998), and non-synaptic membrane specializations, or ³free PSDs´, has been observed using EM analysis in vivo during the initial stages of synaptogenesis (Hinds and Hinds 1976; Blue and Parnavelas 1983; Steward and Falk 1986; Fiala et al. 1998). Taken together, these findings postulate that postsynaptic mechanisms exist to facilitate synapse formation. The findings that postsynaptic molecules such as neuroligins and SynCAM can induce presynaptic assembly, further highlight the importance postsynaptic proteins in driving the formation of new contacts (Scheiffele et al. 2000; Biederer et al. 2002).  Recent advances in imaging techniques have revealed some of the molecular events underlying formation of excitatory synapses in the CNS (Friedman et al. 2000; Sanes and Lichtman 2001; Ahmari and Smith 2002; Washbourne et al. 2002; Li and Sheng 2003; Shapira, Zhai et al. 2003; Bresler et al. 2004; Ziv and Garner 2004; Waites et al. 2005). These elegant studies demonstrated that the major components of the synaptic vesicle release machinery travel in transport packets and are rapidly recruited to contact sites. Other studies revealed transport packets and protein complexes that regulate delivery of NMDA receptors to nascent neuronal contacts (Washbourne et al. 2002; Sans et al. 2003; Washbourne et al. 2004; Waites et al. 2005). The recruitment of NMDA receptor clusters slightly preceded that of synaptic vesicle proteins at newly formed contacts between axonal growth cones and dendrites (Washbourne et al., 2002). Live imaging has also identified discrete and mobile PSD-95 clusters (Marrs et al. 2001; Prange and Murphy 2001).  This suggested the existence of prefabricated PSD-95 clusters potentially involved in the formation and/or stabilization of synapses. However, others reported that presynaptic differentiation precedes recruitment of PSD-95 to newly formed synaptic sites and that PSD-95 accumulation occurs gradually (Friedman et al. 2000; Bresler et al. 2004; Waites et al. 2005).  71  Here we assessed trafficking of several scaffolding proteins in young neurons, prior to the majority of synapse formation. We found that: (1) stationary complex containing PSD- 95, GKAP and Shank are abundant in young neurons at both synaptic and non-synaptic sites, (2) a small fraction of this scaffold complex is mobile, and can be recruited to nascent and existing presynaptic contacts, (3) a subset of stationary protein complex clusters contains NL1 and recruits synaptophysin-positive axonal transport vesicles that can recycle the vital dye FM 4-64, (4) assembly of the scaffold protein complex requires PSD-95, as interfering with PSD-95 expression by siRNA disrupts clustering of GKAP and Shank, and reduces the number of excitatory synapses in young hippocampal neurons, and finally (5) knock down of PSD-95 results in an overall increase in VGAT puncta positive for NL1, indicating a shift in NL1 localization form excitatory to inhibitory contacts.  72 2.2 MATERIALS AND METHODS  Additional and detailed Material and Methods are also found in Chapter 5. 2.2.1 cDNA constructs. Constructs for PSD-95 GFP, PSD-95 YFP, GKAP CFP and GKAP DsRED were constructed by PCR and subcloned in frame into HindIII and EcoRI restriction sites of GW1-GFP, eYFP-C1, eCFP-C1, and pDsRED2-C1 vectors respectively (Clonetech). Shank-YFP and Shank-CFP were generated as previously described (Romorini et al. 2004). Neuroligin-1 CFP was a gift from Dr. P. Washbourne (University of Oregon) and was made as described (Fu et al. 2003). Synaptophysin was constructed by PCR in frame with DsRED into the XhoI and BamHI restriction sites of the pDsRED2-C1 (Clonetech). PSD-95 and scrambled siRNA was subcloned into pSUPER (Clonetech) as previously described (Nakagawa et al. 2004).   2.2.2 Neuronal cell culture and transfection.  Dissociated primary neuronal cultures were prepared from hippocampi of embryonic day 18/19 Wistar rats, and maintained in NeuroBasal media (GIBCO-Invitrogen) supplemented with B-27, penicillin, streptomycin, and L-glutamine. Neurons were plated at 100k per 35 mm glass bottom microwell dish (MatTek). Transfection was preformed using 0.2 µg DNA and 0.2 µl of Lipofectamine-2000 (GIBCO-Invitrogen) for 3 hours. Nucleofection was performed at plating as described by the manufacturer (Amaxa). Briefly, 4 million neurons were suspended, with 4µg of DNA in the nuleofection solution provided. Cells were plated at a final density of 100k and allowed to recover in DMEM with 10% Calf Serum for 1 hour before replacement with supplemented NeuroBasal Media.  For siRNA experiments, plasmids were transfected using calcium phosphate  73 precipitation at DIV2 and stained at DIV8 as previously described (Passafaro, Nakagawa et al. 2003).  2.2.3 Immunocytochemistry  Coverslips were fixed in ±20C methanol and immunolabelled for synaptic proteins. Mouse monoclonal antibodies include: PSD-95 (1:500; Affinity BioReagents, Golden CO), NR1 (1:200; Synaptic Systems), neuroligin-1 (1:500, Synaptic Systems). Rabbit polyclonal antibodies include: GKAP (1:200; gift from C. Sala), synaptophysin (1:1000; PharMingen), GluR1 (1:500; Upstate Biotechnology, Lake Placid NY), VGLUT1 (1:1000; Synaptic Systems). Guinea pig polyclonal antibodies include: GFP (1:300), Shank (1:500; gift from C. Sala). Secondary antibodies were generated in goat, and conjugated with Alexa488 (1:1000), Alexa568 (1:1000), or Alexa360 (1:400; Molecular Probes) All antibody reactions were preformed in blocking solution (2% normal goat serum) for 1 hour at room temperature or overnight at 4 0C.   2.2.4 Assessment of presynaptic terminal function using FM 4-64.  15 µM FM 4-64 (Molecular Probes) was loaded for 30 seconds into presynaptic terminals using a hyperkalemic solution of 90mM KCl2 in modified HBSS, where equimolar NaCl2 was omitted for final osmolality of 310 mOsm. Neurons were rinsed three times and maintained in HBSS without Ca2+ and in the presence of 5 µM Mg2+ to prevent unloading during image acquisition. 1mM ADVESAP-7 (Sigma) was added to quench non-specific signal. A minimum of three images were captured to confirm positive sites FM loading were stationary presynaptic terminals and not orphan sites (Krueger et al. 2003). Unloading was performed for 30 seconds in the same hyperkalemic solution, and washed three times with NeuroBasal media for continued imaging. Unloading of FM 4-64, was observed in most of labeled presynaptic terminals. Puncta that did not significantly  74 unload to 15% of initial intensity were not included in analysis. The fidelity of FM 4-64 to label excitatory presynaptic terminals was tested by immuno-labeling for the excitatory presynaptic marker VGLUT (Appendix 5.4).   2.2.5 Imaging and analysis  Imaging was preformed 24-36 hours post-transfection in an environmentally controlled stage (37 0C and 5% CO2) with an objective heater. Images were acquired on a Zeiss Axiovert M200 motorized microscope with a 63X 1.4 NA ACROMAT oil immersion lens and a monochrome 14-bit Zeiss Axiocam HR charged-coupled camera with 1300 X 1030 pixels.  Exposures were preformed at 1/3 saturation (200-800ms) and binned 2X2 to minimize photo-damage to live cells, and No visible bleed-through or cross-excitation was detectable at 3000ms (Appendix 5.1). To correct for out-of-focus clusters within the field of view, focal plane (z-) stacks were acquired and maximum intensity projections performed offline. Images were scaled to 16 bits and analyzed in Northern Eclipse (Empix Imaging, Missasauga, Canada), by using user-written software routines (Appendix 5.2). Two-tailed parametric Student¶s T-test was preformed to calculate statistical significance of results between experimental groups. All n values represent number of neurons examined from 2-6 independent experiments and are indicated in figure legends. S.E.M. values were calculated based on number of neurons examined.    75 2.3 RESULTS  2.3.1 PSD-95, GKAP, and Shank co-localize at non-synaptic sites in young neurons  Previous studies showed clusters of PSD-95 and GKAP at early stages of neuronal development, and that many of these clusters exist at non-synaptic sites (Rao et al. 1998). We further evaluated the content and location of these clusters at different stages of neuronal development. At day in vitro (DIV) 7, PSD-95 co-localized in discrete clusters along the dendritic shaft with the scaffolding proteins GKAP and Shank (87.6±2.8% of PSD-95 and GKAP co-clusters with Shank; Figure 2.1A). Similar co-localization of these proteins was observed at DIV 14 (86.6±2.3% of PSD-95 and GKAP co-clusters with Shank; Figure 2.1B). Although most of these clusters were synaptic at DIV 14 (78.2±3.7% of PSD-95 and Shank co-clusters with synaptophysin), only half of the clusters containing PSD-95, GKAP and Shank were synaptic at DIV 7 (47.2±5.2% of PSD-95 and Shank co-clusters with synaptophysin, p<0.001, Figure 2.1 C,D). AMPA receptors, as assessed by GluR1 staining, did not significantly co-localize with Shank at DIV 7, and only half of the synaptic Shank clusters contain the glutamate receptor subunit, GluR1 (51.5±6.2% of Shank and synaptophysin co-clusters with GluR1, p<0.001, Figure 2.1 E,F), suggesting that most synapses formed at this early stage are silent (Bredt and Nicoll 2003). The tight correlation in localization of PSD-95, GKAP and Shank at non-synaptic sites at a period that precedes the majority of synapse formation suggests that these proteins form a preassembled complex that participates in excitatory synapse development.       76   Figure 2.1: Detection of non-synaptic clusters of postsynaptic scaffolding proteins. Endogenous localization of synaptic proteins at DIV 7 (left panels) and DIV 14 (right panels) by immuno- staining of primary cultures of hippocampal neurons. (A,B) Scaffolding molecules Shank (green), GKAP (red), and PSD-95 (blue), are co-localized at DIV 7 (87.6±2.8%) and DIV 14 (86.6±2.3%). (C,D) At DIV 7, 47.2±5.2%  of Shank (green) and PSD-95 (red) co-clusters were apposed by synaptophysin (SYN) (blue). At DIV 14, 78.2±3.7%   of PSD-95 and Shank co-clusters are apposed by SYN. (E,F) At DIV 7, only 51.5±6.2% of synaptic sites, containing Shank (green) and SYN (blue), contain the AMPA receptor subunit GluR1 (red). GluR1 clusters are present at 82.4±0.3% of synaptic Shank sites at DIV 14. Data represents analysis of least 2 experiments, n=11-15 neurons, •800 clusters per group. Scale bar, 1µm. ***p<0.001.  77 2.3.2 Mobile and stationary clusters containing PSD-95, GKAP and Shank  To visualize the formation and assembly of clusters of PSD-95, GKAP and Shank we performed time-lapse microscopy of fluorescently-tagged versions of these proteins in DIV5-6 hippocampal neurons using an environmentally controlled chamber, with images taken every 2-10 minutes for periods of up to 2 hours. To avoid any changes associated with protein over-expression, several criteria were followed to ensure that the level and targeting of fluorescently labeled molecules were similar to endogenous proteins (Chapter 5.3). The majority of co-clusters containing PSD-95 GFP and GKAP DsRED were stationary (77.9±2.2% of total co-clusters). Noteworthy, clusters of these proteins were also mobile (10.5±0.9%) or split (12.7±1.3%), moving perfectly in concert along dendrites in both anterograde and retrograde directions (Figure 2.2A).  The mean speed of mobile clusters in a specific direction was 0.83±0.08 µm/min, over distances ranging from 2 to 15 µm (Figure 2.2B). The frequency of mean speeds of mobile clusters for PSD-95 GFP alone and PSD-95 co-transfected with GKAP show similar distribution (Figure 2.2C). Movement was not constant, with clusters often pausing briefly before continuing.   Figure 2.2: Quantification of the mobility of PSD-95, GKAP and Shank. (A) Summary of mobile (10.5±0.9%), splitting (12.7±1.3%), and stationary (77.9±2.2%) co-clusters containing PSD-95 GFP and GKAP DsRED (n=13 neurons, 886 puncta). (B) Mean velocity of mobile PSD-95 and GKAP co-clusters was 0.83±0.08 µm/min, and was not significantly different when PSD-95 alone (0.75±0.03 µm/min, n=21 neurons), or when co-transfected with Shank (0.74±0.08 µm/min, n=10 neurons), or Shank alone (0.77±0.05 µm/min, n=9 neurons, p=0.9). (C) Frequency of mean velocities of mobile clusters for PSD-95 GFP alone (black bars) and PSD-95 co-transfected with GKAP (grey bars) show similar distribution.  78  These characteristics were similar in neurons expressing PSD-95 GFP alone (Figure 2.3 A), co-expressed with GKAP DsRED (Figure 2.3 B), and co-expressed with Shank CFP (Figure 2.3 C). Fusion and splitting of co-clusters of these proteins were observed along the dendrite, with modest, less than 2µm, lateral movement (12.7±1.3% of total co- clusters, Figure 2.2 D).    Figure 2.3: Visualization of mobile and stationary clusters containing PSD-95, GKAP and Shank. Time-lapse imaging of hippocampal neurons transfected at DIV 5-6 with PSD-95 GFP alone, PSD-95 GFP and GKAP DsRED, or with PSD-95 YFP and Shank CFP. Open arrow heads point to mobile clusters of interest, and closed arrowheads demarcate their original position. A small fraction of clusters of (A) PSD- 95 GFP alone or with (B) GKAP DsRED move in anterograde and retrograde direction along the dendrites. (C) Mobile co-clusters of PSD-95 YFP and Shank CFP were also observed (D) Co-clusters of PSD-95 GFP and GKAP DsRED were observed to split and merge with other existing clusters. Scale bar, 10µm.   79 This movement is consistent with previous reports of PSD-95 alone at this age (Friedman et al. 2000; Bresler, Ramati et al. 2001; Washbourne et al. 2002; Shapira et al. 2003; Bresler et al. 2004), albeit lower than those typically associated with vesicles propelled by molecular motors (0.1-10 µm/sec), and faster than previous reports of PSD-95 alone in older neurons (0.48 µm/min (Bresler et al. 2001). Thus, these data indicate that different pools of non-synaptic scaffold complexes exist, which are distinguishable by differences in mobility. The detection of both stationary and mobile clusters containing PSD-95, GKAP and Shank is intriguing, and suggests that these proteins constitute a complex that can move as a bonafide transport packet. Since the protein clusters described here contain the major scaffolding proteins present at excitatory synapses, we will refer to them as the preformed scaffold complex.   2.3.3 Stationary non-synaptic preformed scaffold complexes participate in development of functional presynaptic terminals  At DIV 5, the number of endogenous PSD-95 puncta was about 2 fold higher than sites positive for the excitatory presynaptic marker, vesicular glutamate transporter-1 (VGLUT).  By DIV 9, there was an increase in the total number of co-localized VGLUT and PSD-95 puncta, without a significant increase in the total number of PSD-95 clusters (Figure 2.3), suggesting that existing PSD-95 clusters may have been utilized for building new contacts positive for VGLUT within this time period.   80  Figure 2.4: Co-localization of PSD-95 and VGlut increases with developmental age. (A) Representative images showing changes in number of endogenous PSD-95 and VGLUT clusters at DIV 5 (left panels) and DIV 9 (right panels) neurons. (B) The number of PSD-95 per unit length did not change significantly from DIV 5 to 9 (102±3%, p=0.8). The number of VGLUT positive sites increased 150±7% (p=0.06), as well as the number of sites positive for VGLUT and PSD-95 (232±24%, p<0.001). Scale bar, 5µm.   To explore this possibility, time-lapse imaging of PSD-95 GFP was performed in conjunction with successive loading and unloading of the synaptic vital dye FM 4-64 (Prange and M urphy 1999; Friedman et al. 2000; Shapira et al. 2003). Sites positive for FM 4-64 were also labeled for VGLUT (83.6±19% of total FM 4-64 positive sites), confirming the validity of using FM 4-64 to label excitatory presynaptic terminals (Appendix 5.4). When compared to their synaptic, FM 4-64 positive counterparts over a time-course of 4 hours, there was no statistically significant difference in relative fluorescence intensity of PSD-95 GFP: Final GFP fluorescence intensity was 90.2±1.6% of initial fluorescence intensity for non-synaptic PSD-95 GFP, and 92.9±4.1% of initial fluorescence intensity for synaptic PSD-95 GFP (n=10 neurons per group, 378 puncta, p=0.6). The fluorescence stability of stationary PSD-95 GFP clusters negative for FM 4- 64 suggests that they are not simply remnants of recently lost synapses, which have been shown to disassemble within 1 hour (Okabe, Miwa et al. 2001).  Within 45 minutes of imaging, a small fraction of presynaptically naive sites opposed to stable PSD-95 GFP clusters became FM 4-64 positive (4.5±1.4% of total PSD-95 puncta, 10.4±3.7% of presynaptically naïve PSD-95 puncta). This was observed in 6/12 of the neurons analyzed (Figure 2.5A). After 2 hours of imaging, this phenomenon was  81 observed more frequently (11.3±2.0% of total PSD-95 puncta, 18.6±3.4% of presynaptically naïve PSD-95 puncta), in 13/14 neurons examined (Figure 2.5B). These data suggest that preformed scaffold complexes can act as sites for recruitment of active presynaptic machinery, and that this process can occur in as little as 45 minutes.   Figure 2.5: Sites apposed to stationary non-synaptic scaffold clusters are readily transformed to active presynaptic terminals. (A,B) DIV 5 hippocampal neurons were transfected with PSD-95 GFP and analyzed 24-36 hours post- transfection for changes in number of active presynaptic terminals by subsequent loading (left panels) and unloading (right panels) of FM 4-64. Examples of stationary FM 4-64 negative PSD-95 GFP (closed arrowhead, load 1), which became FM 4-64 positive within 45 minutes (A,A¶) and 2 hours (B,B¶) of initial load (load 2, n=12,14 neurons).   2.3.4 Delivery of mobile synaptophysin clusters to contact sites apposed to stationary preformed scaffold complexes  Although FM dyes are reliable markers for labeling active presynaptic terminals, this method does not readily allow for visualization of delivery of presynaptic elements to these sites. To visualize axonal transport packets recruitment at contact sites we used synaptophysin fused to DsRED (SYN DsRED) introduced at DIV 0 by electroporation in  82 order to label a large proportion of presynaptic terminals and their axons, followed by transfection of PSD-95 GFP as described earlier. Using fast interframe intervals (15 seconds), we found that mobile SYN DsRED travel on average velocities of 0.47 µm/s (ranging from 0.14 to 1.3 µm/s; Figure 2.6 A,B) similar to  other presynaptic vesicle markers such as VAMP (Ahmari et al 2000). Frequent rests in trafficking of SYN DsRED positive transport packets averaged 54 seconds (ranging from 15 sec to 6.5 min;. Figure 2.6C). Recruitment of SYN DsRED positive transport packets at specific sites for a minimum of 15 minutes was counted as stable in order to distinguish it from these rest stops. The level of overexpression of SYN DsRED was minimal and overlapped with endogenous synaptophysin (Appendix 5.3). This data represents a reconstruction and confirmation of the published data with a new SYN DsRED construct.   Figure 2.6: Characterization of transport packets positive for synaptophysin DsRED. (A) Yellow arrow head demarcates a stable SYN DsRED cluster. Red and green arrow heads follow mobile clusters (B) Frequency distribution of the average velocity of SYN DsRED transport packets. (C) Frequency distribution of the stop time of SYN DsRED transport packets. Data represents 3 experiments, 6 fields of view, >200 axons. Scale bar (A,B), 5µm.  83 Time-lapse revealed 36.3±5.4% of PSD-95 GFP clusters contacting SYN DsRED- positive axons showed stable PSD-95 GFP clusters apposed by stable accumulations of SYN DsRED, suggesting that these are synaptic partners. Several contacts between PSD- 95 GFP puncta apposed by SYN DsRED-positive axons showed no axonal varicosities (bulges at least 1.5 times greater than average width of the axon) and no significant enrichment of SYN DsRED (increases in fluorescence at least 1.5 times background).  A number of these contacts were seen to develop into axonal varicosities (Figure 2.7A) or recruit and stabilize mobile axonal transport packets positive for SYN DsRED (Figure 2.7B, from re-constructed SYN DsRED). Within 1 hour 5.9±1.7% of total PSD-95 GFP clusters (5/88), or 15.6±1.4% of presynaptically naïve PSD-95 GFP clusters (5/31) recruited SYN DsRED (Figure 2.7 C,D). This data has also been recapitulated with the re-constructed SYN DsRED, where 2/23 presynaptically naïve PSD-95 GFP clusters recruited SYN DsRED (n=12 neurons, 2 independent experiments).  Some of the axons showed rapid morphological re-arrangements upon delivery of SYN DsRED positive transport packets to sites apposed to stationary PSD-95 clusters, suggesting that these changes are triggered by crosstalk between these pre- and postsynaptic elements (Figure 2.7A). Longer time-lapse periods (2 hours) revealed that 32.5±4.6% of presynaptically naïve contacts positive for PSD-95 GFP (11/32) were able to recruit SYN DsRED (Figure 2.7 C,D). These results indicate that stationary clusters of preformed postsynaptic scaffolds can recruit axonal transport packets for initiation and/or stabilization of new sites of contact, and suggests a paradigm whereby preformed postsynaptic complexes may induce presynaptic differentiation.  Previous studies have indicated that differentiation of presynaptic compartment occurs before recruitment of postsynaptic proteins (Bresler et al. 2001; Okabe et al. 2001; Waites et al. 2005). To further assess whether both of these modes exist in DIV5-7 neurons, we monitored accumulation of PSD-95 GFP at sites initially lacking PSD-95 but positive for SYN DsRED.  Within 2 hours of imaging, 26.7±3.6% of stable SYN DsRED clusters recruited PSD-95 GFP at sites initially naïve for this postsynaptic protein (5/25; Figure 2.7D). Three of these events resulted from recruitment of discrete PSD-95 GFP clusters that were in the vicinity of the contact. In two other events no discernable  84 clusters were recruited and may represent the accumulation of a diffuse pool as previously described (Bresler et al. 2001; Okabe et al. 2001; Washbourne et al. 2002; Waites et al. 2005).  Thus, both pre- and postsynaptic mechanisms participate in the recruitment of synaptic elements at sites of contact. In contrast with these findings, only 2.5±1.4% of points where PSD-95GFP transfected dendrites crossed SYN DsRED axons, and were initially negative for both PSD-95 GFP and SYN DsRED, were observed to recruit and stabilize SYN DsRED within 2 hours (1/41; Figure 2.7E).  Thus, a contact per se is not sufficient to drive synapse formation.  These results emphasize the importance of the preformed scaffold complex in increasing the probability of accumulation of presynaptic elements.    Figure 2.7: Time-lapse of SYN DsRED and PSD-95 GFP. (A,B) Two examples of presynaptically naïve PSD-95 GFP clusters (closed arrowheads) are shown to recruit SYN DsRED clusters, which remained as stable apposed pairs for duration of the imaging period. (C)  Percentage of PSD-95 GFP puncta, including both those initially negative and positive for SYN DsRED clusters or (D) those only negative (presynaptically naïve) for SYN DsRED clusters, that recruit SYN DsRED transport packets within 1 hour (n=20 neurons), or 2 hours (n=11 neurons) of imaging. (E) Comparison of different modes of recruitment of PSD-95 GFP and SYN DsRED occurred within the imaging period. Within 2 hours of imaging, 32.5±4.6% of presynaptically naïve contacts positive for PSD-  85 95 GFP recruited SYN DsRED, whereas 26.7±3.6% of stable SYN DsRED clusters recruited PSD-95 GFP at sites initially naïve for this postsynaptic protein.  Only 2.5±1.4% of contacts initially negative for both PSD-95 GFP and SYN DsRED were observed to recruit and stabilize SYN DsRED (n=11). Scale bar, 1µm.   Accumulation of pre- and postsynaptic proteins was mainly observed at existing contacts between transfected neurons; thus, it remains unclear how postsynaptic clusters attract axons to these sites. We have explored this by searching for events in which axonal growth cones contacted dendrites of transfected cells. We were able to capture a total of six growth cones contacting dendrites of cells transfected with PSD-95 GFP, and during five of these events, axonal contacts were maintained with existing PSD-95 GFP clusters (Figure 2.8).   Figure 2.8: Dynamics of a growth cone contacting PSD-95 GFP clusters. Representative examples of a DsRED positive axonal growth cone that contacted PSD-95 GFP clusters (demarcated by arrowhead), and maintained contact for the duration of the imaging period.  86 2.3.5 Stationary non-synaptic preformed scaffold complexes contain neuroligin-1  To define signals that contribute to the stability of preformed scaffold complex, we analyzed the localization of NL1, a binding partner of PSD-95 that is sufficient to induce presynaptic differentiation (Scheiffele et al. 2000). In DIV7 neurons, 56.5±0.5%  of Shank clusters are positive for NL1 (Figure 2.9 A,B).  However, only 28.1±5.3% of these clusters were associated with excitatory presynaptic terminals, as assessed by VGLUT staining (Figure 2.9 C,D). Thus, NL1 can be associated with the preformed complex at synaptic and non-synaptic sites. At DIV14 in contrast, NL1 clusters mainly exist with scaffold protein complexes (62.2±6.5%) and synaptophysin positive sites (64.8±4.7%).    Figure 2.9: Neuroligin-1 is associated with Synaptic and non-synaptic preformed scaffold clusters. Representative images of localization of endogenous neuroligin-1 (NLG1) with respect to postsynaptic and presynaptic proteins as assessed by immunostaining at DIV 7 (left panels) and DIV 14 (right panels). (A-D) Localization of NLG1 and Shank clusters with (A,B) GKAP and (C,D) VGLUT  at DIV 7 and DIV 14 (n=11 neurons at DIV 7; 12 neurons at DIV 14; •1000 clusters). Scale bar, (A,C) 1µm, ***p<0.001.   87 To further dissect the pool of preformed protein complexes containing NL1, time-lapse imaging of PSD-95 GFP and GKAP DsRED transfected neurons was performed, followed by retrospective immunostaining for endogenous NL1. NL1 clusters were detected in most stationary PSD-95 and GKAP co-clusters (75.1±3.1% of total stationary PSD-95 clusters). A fraction of the mobile PSD-95/GKAP co-clusters were found associated with NL1 puncta (33.3±11.4% of mobile PSD-95 clusters; Figure 2.10 A,B), representing ~3% of the total number of PSD-95 clusters, and moved distances less than 2.5 µm.    Figure 2.10: Stationary clusters contain neuroligin-1 (A) Open arrowhead highlights a mobile cluster of PSD-95 GFP and GKAP DsRED (left panels) that is negative for endogenous NLG1 (right panels). (B) Endogenous NLG1 is present at 75.1±3.1% of stationary PSD-95 GFP and GKAP DsRED co-clusters. In contrast, 33.3±11.4% of mobile PSD-95 GFP and GKAP DsRED co-clusters contain NLG1 (n=12 neurons; 321 puncta). (C) Clusters of PSD-95 GFP negative for FM 4-64 and SYN contain NLG1 (closed arrowheads).  (D) Percentage of PSD-95 GFP clusters positive for NLG1 at sites positive (synaptic) or negative (non-synaptic) for both FM 4-64 and SYN (n=12 neurons, 299 puncta, p=0.7). Scale bar, (A,C) 5µm. ***p<0.001.  88 To determine the relationship between NL1, stationary PSD-95 clusters and active presynaptic terminals, neurons expressing PSD-95 GFP were loaded and unloaded with FM 4-64, time-lapse imaging was performed over a 1-4 hour period, and neurons were fixed and stained for endogenous NL1 and synaptophysin. NL1 was found associated with stationary PSD-95 clusters, regardless of their apposition to active presynaptic terminals. (58.9±8.6% of total FM 4-64 and synaptophysin positive clusters versus 53.2±9.0% of total negative clusters, p=0.7; Figure 2.10 C,D).  Consistent with these findings, time-lapse imaging of DIV 6 neurons expressing PSD-95 YFP and NL1 CFP showed a large proportion of the PSD-95 YFP clusters were stationary during the imaging period (88.7±2.5% of total PSD-95 clusters, Figure 2.11 A), and that most of these stationary PSD-95 YFP clusters contained NL1 CFP (89.7±2.5%; Figure 2.11 C). Most mobile PSD-95 YFP clusters did not contain NL1 CFP (Figure 2.11 B), however, similar to retrospective staining for endogenous neuroligin-1, a small fraction of the mobile pool of PSD-95 YFP clusters at this age contained NL1 CFP (27.3±3.9% of mobile PSD-95 clusters; Figure 2.10C). Interestingly, the average velocity of the few mobile PSD-95 YFP clusters positive for NL1 CFP was significantly slower than those that were negative for NL1 (0.48±0.06 versus 0.84±0.06 µm/min respectively, p<0.001; Figure 2.11 D), and moved shorter distances (on average 2.2 µm versus 6.0 µm respectively; Figure 2.11 E). Thus, with retrospective immunostaining and live time-lapse imaging, a large proportion of NL1 is found associated with stationary PSD-95 clusters.   89   Figure 2.11: Stationary scaffold clusters contain neuroligin-1. (A) Stationary co-clusters of PSD-95 YFP and NLG1 CFP represent 87.8±2.5% of total PSD-95 YFP clusters. (B) A mobile cluster of PSD-95 YFP that is negative for NLG1 CFP (open arrowhead highlights the mobile cluster, closed arrowhead demarcates original position). (C) NLG1 CFP was present at 89.7±2.5% of stationary PSD-95 YFP, 63.5±15.2 of splitting PSD-95 YFP clusters, and 27.3±3.9% of mobile PSD-95 YFP clusters. (D) Average velocity of mobile PSD-95 YFP clusters that contain NLG1 CFP (0.48±0.06 µm/min), versus NLG1-negative clusters (0.84±0.06 µm/min, p<0.001). (E) Frequency distribution of distance traveled by PSD-95 YFP clusters alone (black bars), or with NLG1 CFP (white bars, n=10 neurons). Scale bar, (A,B) 1µm. ***p<0.001.    We next examined whether non-synaptic clusters of NL1 recruits presynaptic proteins. Indeed, stationary clusters of NL1 CFP were observed to recruit axonal transport packets positive for SYN DsRED at sites of contact within 1 hour (16.6±3.6%  of total contacts (13/67 contacts), 47.7±7.9% of SYN DsRED naïve contacts (13/27contacts; Figure 2.12 A). Retrospective immunostaining revealed that 80% (16/20 events) of stationary clusters of NL1 CFP that recruited SYN DsRED contain endogenous PSD-95 (Figure 2.12 B). These data suggest that association of NL1 with the preformed scaffold complex may be one of the adhesion molecules that facilitate recruitment of the presynaptic release machinery.    90   Figure 2.12: Stationary scaffold clusters containing neuroligin-1 recruits synaptophysin positive transport packets. (A) Images of a cluster of NLG1 CFP apposed by synaptophysin DsRED (SYN DsRED) positive axon (closed arrow) which rapidly recruited and stabilized a SYN DsRED transport packet (open arrow). This was observed in 47.7±7.9% of contacts initially negative for SYN DsRED clusters (n=13 neurons). (B) A NLG1 CFP cluster that recruited SYN DsRED transport packet within 75 minutes of imaging, is positive for endogenous PSD-95.  Quantitative analysis shows that 80% of NLG1 CFP clusters that recruit SYN DsRED transport packets are positive for PSD-95 (16/20 events, n=15 neurons). Scale bar, (A) 5 µm, (B) 1µm ***p<0.001.   2.3.6 Mobility of preformed scaffold clusters is actin dependent  If stationary preformed scaffold clusters act as signals for the recruitment of active presynaptic terminals, what is the role of the mobile clusters? Transport of these preassembled complexes is slower than reported values for known vesicular transport. As such, movement of PSD-95 and GKAP clusters appears to involve a novel mechanism of transport. To determine the cytoskeletal elements required for transport of mobile PSD- 95 clusters, time-lapse imaging of transfected neurons was performed in the presence of 4µM cytochalasin B or 3mM nocodozole, pharmacological agents known to disrupt actin and microtubule based transport respectively, through filament depolymerization (Sabo and McAllister 2003).   91   Figure 2.13: PSD-95 movement is actin dependent. Time-lapse imaging of PSD-95 GFP was performed in the presence of nocodozole or cytochalasin B to assess changes in velocity of PSD-95 GFP clusters. (A) Addition of 3 µM nocodozole (+NOCOD) did not significantly alter mobility of PSD-95 GFP clusters (open arrowhead highlights a mobile cluster, closed arrowhead demarcates initial position). (B) Addition of 4 µM cytochalasin B (+CYTOB) immobilized PSD-95 GFP clusters (open arrowhead). (C) Average velocity before nocodozole treatment was 0.78±0.03 µm/min, and (D) 0.69±0.05 µm/min during treatment (p=0.14, n=10 neurons). Average velocity before cytochalasin B treatment was 0.75±0.06 µm/min, and 0.10±0.07 µm/min during treatment (p<0.001, n=9 neurons). Scale bar, 5µm. ***p<0.001.   The relative velocities of mobile clusters were assessed for 30 minutes before and 30 minutes to 1 hour after drug treatment. Movement of PSD-95 GFP clusters was abolished upon treatment with cytochalasin B (0.75±0.06 µm/min before versus 0.10±0.07 µm/min during cytochalasin B treatment, p<0.001; Figure 2.13 B,D), whereas no significant effect was seen with nocodozole (0.78±0.03 µm/min before versus 0.69±0.05 µm/min after nocodozole treatment, p=0.14; Figure 2.13 A,C). These experiments suggest that the mobility of preformed scaffold complexes involves an actin-based trafficking.      92 2.3.7 Mobile preformed scaffold clusters are recruited to nascent and existing postsynaptic sites  Do mobile clusters contribute to synapse formation or remodeling? Simultaneous time- lapse imaging of PSD-95 YFP, GKAP CFP, and FM 4-64 revealed that presynaptic terminals labeled with FM 4-64 recruited co-clusters of PSD-95 YFP and GKAP CFP to nascent and existing postsynaptic sites. Recruitment of PSD-95 YFP and GKAP CFP co- clusters to existing puncta of postsynaptic proteins occurred more frequently (1.8±0.6% of total PSD-95 puncta, 30.1±13.0% of mobile PSD-95 puncta) and was observed in 6/10 neurons during the 1 hour time-lapse period (Figure 2.14 A,C). Recruitment of PSD-95 YFP and GKAP CFP co-clusters to nascent presynaptic terminals was a more rare event (0.6±0.3% of total PSD-95 puncta, 10.0±5.1% of mobile PSD-95 puncta), observed in 4/10 of neurons (Figure 2.14 B,C). These data demonstrate that recruitment of preformed scaffold complexes to both existing and nascent presynaptic terminals can occur through the transport of discrete clusters. Further studies will be required to determine whether these clusters are stably recruited for longer time periods, and whether this recruitment represencts a chance encounter stabilized by protein interactions or these slusters are actively being targeted to these sites. Analysis of FM 4-64 positive versus FM 4-64 negative co-clusters revealed differences in their behavior. A significantly larger fraction of non-synaptic clusters were mobile during the imaging period (2.3±1.2% of total FM 4- 64 positive PSD-95 and GKAP co-clusters versus 17.6±3.0% of total FM 4-64 negative PSD-95 and GKAP co-clusters, p<0.001). Together these data demonstrate that mobile preformed clusters are more likely to be non-synaptic, and can be utilized as transport packets for formation of a PSD at nascent and existing synapses.  93  Figure 2.14: Mobile preformed scaffold clusters are recruited to nascent and existing sites. Time-lapse imaging of PSD-95 YFP and GKAP CFP in conjunction with FM 4-64. (A) A mobile co-cluster of PSD-95 YFP and GKAP CFP (open arrowhead highlights a mobile cluster, closed arrowhead demarcates initial position) recruited to a site positive for FM 4-64 and PSD-95 YFP/GKAP CFP  (yellow arrowhead). (B) An example of a mobile cluster containing PSD-95 YFP and GKAP CFP (open arrowhead highlights a mobile cluster, closed arrowhead demarcates initial position) recruited to a site positive for FM 4-64 but negative for PSD-95 YFP/GKAP CFP  (yellow arrowhead). (C) Quantitative analysis of mobile clusters expressed as percentage of as percentage of total mobile and stationary PSD-95 YFP and GKAP CFP co- clusters (left panel) or as percentage of mobile clusters alone (right panel) recruited to FM4-64 positive (Existing) and negative (Nascent) sites. (D) Graph of the various pools of PSD-95 and GKAP co-clusters (mobile, split and stationary) and their recruitment to FM 4-64 positive and negative sites. n=10 neurons, 448 puncta. Scale bar, 1µm. *** p<0.001.  94 2.3.8 Knock Down of PSD-95 results in a decrease in GKAP, Shank and VGLUT clusters, and an increase in VGAT clusters positive for neuroligin-1  To determine whether PSD-95 is central to the nucleation of the preformed scaffold complexes, siRNA that has previously been shown to knock down PSD-95 expression was transfected at DIV 2, and the effects were assessed at DIV 7-8 (Nakagawa et al. 2004). A significant decrease in number of PSD-95 clusters was observed, from 8.2±0.6 puncta/µm in neurons transfected with scrambled siRNA, to 2.2±0.3 puncta/µm in neurons transfected with PSD-95 siRNA (Figure 2.15). Knock down of PSD-95 also resulted in a significant reduction in number of GKAP (4.3±0.4) and Shank (3.1±0.4) clusters when compared to neurons transfected with scrambled siRNA (7.2±0.5 and 8.4±0.6 puncta/10µm respectively, Figure 2.15 A,B,E). We next examined whether a reduction in the number of clusters of PSD-95, GKAP and Shank is associated with a change in number of the excitatory presynaptic marker, VGLUT. Indeed, a significant decrease in the number of VGLUT-positive puncta contacting neurons transfected with PSD-95 siRNA was observed (1.1±0.2 puncta/10µm) versus scrambled siRNA (2.6±0.5 puncta/10µm; Figure 2.15 C,G).  In contrast with these findings, the total number of NL1 clusters was not significantly altered in neurons transfected with either scrambled (9.0±0.5 puncta/10µm) or PSD-95 siRNA (10.3±0.6 puncta/10µm). Consistent with this, the number of NL1 clusters positive for Shank decreased in PSD-95 siRNA transfected neurons (Figure 2.15 F). This paralleled with a 169±19.1% increase in number of VGAT positive puncta in PSD-95 siRNA transfected cells (44.6±4.4 puncta/neuron versus 26.4±2.7 in scrambled siRNA control neurons; Figure 8H).  This suggests that loss of the preformed scaffold complexes may have resulted in miss-localization of NL1, potentially to inhibitory contacts. To address this, we first analyzed the proportion of VGAT positive contacts that contain NL1 in DIV7 neurons, and found that 27.9±6.2% of VGAT positive sites are apposed by NL1. Next, we contrasted the percentage of NL1 puncta present at VGAT positive sites in PSD-95 siRNA transfected neurons versus controls.  We found that the proportion of  95 NL1 puncta positive for VGAT increased by 150±14.5% in PSD-95 siRNA transfected cells (Figure 2.15 I). These results indicate that preformed scaffold complexes may serve a role in maintaining the balance between newly formed excitatory and inhibitory synapses.     Figure 2.15: Knock-down of PSD-95 reduces clustering of GKAP, Shank, and VGLUT. Neurons transfected with PSD-95 siRNA (right panels) or scrambled siRNA (left panels) at DIV 2 and stained at DIV 7-8 for (A) GKAP and PSD-95, (B) Shank and PSD-95, (C) Shank and VGLUT or (D) NLG1 and VGAT. (E) A significant reduction in the number of puncta/10µm of PSD-95 (n=15), GKAP (n=11), and Shank (n=13). (F) Reduced number of co-clusters positive for NLG1 and Shank in PSD-95 siRNA transfected cells (n=30). (G,H) An increase in number of (G) VGLUT, but a decrease in number of (H) VGAT positive contacts apposed to neurons transfected with PSD-95 siRNA. (I) Number of VGAT and NLG1 co-clusters increased in PSD-95 siRNA transfected cells (n=10 neurons).  Scale bar, 5µm. *p<0.05, **p<0.01, *** p<0.001.  96  2.4 DISCUSSION  Recent studies have shown that synapse formation involves rapid delivery of transport packets containing presynaptic proteins to new sites of contact (Craig and Boudin 2001; Sanes and Lichtman 2001; Ahmari and Smith 2002; Garner et al. 2002; Li and Sheng 2003; Sans et al. 2003; Waites et al. 2005). Others have documented non-synaptic postsynaptic protein complexes (Blue and Parnavelas 1983; Fiala et al. 1998; Rao et al. 1998; Washbourne et al. 2002; Sans et al. 2003). Here we show that mobile and stationary preformed protein complexes containing PSD-95, GKAP and Shank, can participate in excitatory synapse development via recruitment of presynaptic transport packets positive for synaptophysin. A significant proportion of these stationary non- synaptic clusters contain NL1, and are readily transformed into FM 4-64 positive axonal terminals, suggesting that the preformed postsynaptic protein complexes assist in transformation of these sites to active presynaptic contacts. These results postulate a mechanism whereby stationary preformed clusters of postsynaptic proteins predetermine the sites at which excitatory synapses are formed.  The presence of PSDs without corresponding functional presynaptic terminals have been observed by EM in vivo (Hinds and Hinds 1976; Blue and Parnavelas 1983; Steward and Falk 1986; Fiala et al. 1998) however, it was unknown whether these ³free PSDs´ could eventually recruit presynaptic release machinery.  Our analysis shows that postsynaptic differentiation can occur prior to formation of a functional presynaptic active zone in young hippocampal neurons. This is consistent with previous findings which demonstrated the existence of postsynaptic protein complexes containing NMDA receptors (Washbourne et al. 2002; Sans et al. 2003; Washbourne et al. 2004).  Live imaging of GFP-tagged NMDA receptor subunit NR1 in young neurons also revealed a modular transport of NR1 and SAP-102 along microtubules, and that NR1 recruitment at new contacts slightly precedes or follow the recruitment of presynaptic proteins (Washbourne et al. 2002; Sans et al. 2003; Washbourne et al. 2004). These NR1 transport packets were lacking PSD-95, indicating that delivery of NMDA receptor subunits to  97 synaptic sites involves a mechanism independent of recruitment of the preformed scaffold complexes examined in this study. The availability of a readily accessible pool of preformed scaffold complexes could facilitate synapse formation at this early stage of development. However, the involvement of preformed postsynaptic complexes involved in synapse formation has been disputed by others who reported that presynaptic differentiation precedes the recruitment of postsynaptic proteins which occurs gradually (Bresler et al. 2001; Marrs et al. 2001; Okabe et al. 2001; Bresler et al. 2004). These studies were performed in relatively older neurons and may have therefore failed to observe some of the events that occur in younger neurons. A shift in mechanisms used for synapse assembly at different developmental stages may explain these discrepancies.  We were able to capture both pre- and postsynaptic mechanisms of recruitment of synaptic proteins to nascent contacts in DIV 5-7 neurons.  Analysis of synaptophysin DsRED positive axons contacting dendrites of cells transfected with PSD-95 GFP showed that ~33% of contacts positive for the postsynaptic scaffold clusters but initially naïve for presynaptic elements, recruited synaptophysin within two hours. Moreover, ~27% of contacts initially naïve for PSD-95 clusters, recruited PSD-95 GFP.  In contrast with these observations, only ~2.5% of sites lacking both PSD-95 GFP and synaptophysin clusters resulted in the accumulation of presynaptic elements within 2 hours, indicating that a contact between two neurons per se is not sufficient to drive synapse formation. This further highlights the importance of the preformed scaffold complexes in triggering accumulation of synaptic elements at these sites.  Most of the new synapses observed in our analysis were en passant, and frequently observed changes in axon morphology accompanied by accumulation of synaptophysin at sites contacting PSD-95 clusters. Events where axonal growth cones made a contact with PSD-95 clusters was rare, however, most events in which axonal growth cones contacted PSD-95 clusters resulted in association of the axon with the PSD-95 cluster until the completion of the time-lapse experiment (at least 30 minutes). It remains unclear why axonal growth cones were initially attracted to the postsynaptic clusters. Secreted factors associated with the postsynaptic scaffold may attract extending axons; however there is  98 no evidence in the current study to support this possibility. It is possible that during their growth, extending axons randomly contact multiple sites, however once they encounter a postsynaptic protein complex containing the appropriate adhesion molecules, they can be rapidly stabilized.  The presence of NL1 in the preformed complex may facilitate initial contact stabilization, however, our data does not exclude the involvement of other cell adhesion molecules in this process.  There are inherent advantages to both the presynaptic and postsynaptic modes of synapse formation for proper connectivity of a neuronal circuit. In early development, guidance cues drives an axon into its respective target field, contacts the correct neuron, and thus presynaptic mechanisms may ensure recruitment of the appropriate scaffold and receptors that matches the neurotransmitter present in axonal terminals.  Conversely, a dendrite primed with the appropriate postsynaptic complex may serve to determine the number of sites to be stabilized or eliminated upon encountering of an axon en passant, and this may dictate the number and type of synapses a neuron receives. Considering that the preformed pool is scarce in older neurons, this postsynaptic mechanism may not significantly contribute to synapse formation in more mature neurons, and must rely on synapse addition by presynaptic induction on a point by point basis. Experiments studying the adhesion complex formed by neuroligins and neurexins suggest that both presynaptic and postsynaptic mechanisms stimulate synapse assembly: When expressed in non-neuronal cells, β-neurexin is sufficient to drive the recruitment of postsynaptic proteins, and conversely neuroligins are sufficient to drive the recruitment of the presynaptic release machinery (Levinson et al. 2005). Studies suggest that scaffolding proteins play an important role in dictating the behavior of cell adhesion molecules in synapse development. For instance, PSD-95 enhances neuroligin-1 clustering and maturation of excitatory synapses at the expense of inhibitory contacts (Prange et al. 2004).  Thus, factors that govern early assembly of these proteins at non-synaptic sites may be critical for controlling the number of newly formed synapses.   99 We analyzed a subset of postsynaptic proteins present at non-synaptic clusters; however it is possible that this complex may contain many other scaffolding proteins and adhesion molecules. The association of PSD-95 and GKAP in young hippocampal neurons in non- synaptic clusters has been previously observed (Rao et al. 1998), however the finding that Shank is also present is intriguing. Shank is a large protein with multiple motifs for protein-protein interaction. Hence, a preformed complex containing PSD-95, GKAP and Shank may be able to recruit many proteins required for excitatory synapse maturation. This hypothesis is supported by the fact that Shank and GKAP form aggresomes which are degraded via the proteosomal pathway in the absence of PSD-95 (Romorini et al. 2004). Shank is functionally involved in the morphogenesis of spines and requires the interaction of Shank with Homer, a protein that binds metabotropic glutamate receptors, and inositol 1,4,5-triphosphate receptors (Tu et al. 1999; Sala et al. 2001). The synapses of young hippocampal neurons observed in this study are predominantly on the shafts of dendrites. Thus, the presence of Shank in the preformed postsynaptic complex suggests involvement of shaft synapses in the later development of spines.  Although less numerous than their stable counterparts, mobile non-synaptic clusters of PSD-95, GKAP and Shank were observed. Previous studies have shown some movement of PSD-95 clusters, including splitting, lateral movement in shafts, and movement into and out of spines and filopodia (Bresler, Ramati et al. 2001; Marrs, Green et al. 2001; Prange and Murphy 2001; Bresler, Shapira et al. 2004). These studies did not address the relationship of these moving clusters with respect to other scaffolding proteins, adhesion molecules, or active presynaptic terminals. In this study, we found mobile clusters of multiple scaffolding proteins are recruited to nascent and existing synapses. The existence of large preformed clusters of postsynaptic proteins that frequently split suggests that these clusters may serve as a reservoir for rapid delivery of preassembled complexes to nascent and established sites. Most intriguing is the finding that the majority of mobile PSD-95 clusters containing NL1 significantly moved slower and for distances less than 2.5 µm. The association of NL1 with a small pool of mobile clusters that exhibit slow movement and travel for short distances suggests that NL1 and PSD-95 may initially traffic separately, but once they merge, NL1 restricts the mobility of the  100 preformed complexes and promote their docking at specific hot spots to prime them to associate with newly encountered axons arriving from excitatory neurons. Our analysis cannot rule out the possibility that individual neuroligin molecules rather than clusters associate with the preformed scaffold complex during new contact formation, mainly because of the methods used here which only visualized clusters of these proteins. Future experiments that can track movement of individual neuroligin molecules will be important to clarify this issue.  Signals that dictate the splitting, transport and docking of preformed clusters at specific subcellular locations remain unclear. EM studies showed association of PSD-95 in vesiculotubular structures that resemble endosomes (El-Husseini et al. 2000). Other studies revealed a fraction of PSD-95 is associated with, yet unidentified, intracellular membranes (Bresler et al. 2001). Thus, mobile clusters may represent a form of endosomal structures, however with a speed distinct from than previously reported. Our analysis suggests that mobility of preformed postsynaptic clusters is actin-dependent. Actin-based trafficking in hippocampal neurons has been shown to be carried out by the myosin family of motor proteins, and several members are expressed in the dendrites of hippocampal neurons (Kim and Sheng 2004). Myosin 5a is a likely candidate since it is highly expressed in young hippocampal neurons, can be found at non-synaptic sites containing PSD-95, and interacts with GKAP through dyenin light chain (Naisbitt, Valtschanoff et al. 2000).  Prange et. al, (2004) analyzed changes in synapses upon over-expression of PSD-95 and NL1 and showed that PSD-95 enhanced accumulation of NL1 at excitatory synapses, thus limiting the number of new synapses induced by NL1.  The current study sheds more light into the potential mechanism that involves assembly of these proteins at neuronal contacts with near-physiological levels of expression. Preformed scaffold complexes may act to control the function of NL by sequestering it at ³hot-spots´, eventually leading to recruitment of presynaptic release machinery. Thus, the number of existing ³hot-spots´ may be critical for controlling the balance of excitatory and inhibitory contacts. Consistent with this role, sequestration of specific NLs to either excitatory and inhibitory  101 contacts has been recently reported (Graf, Zhang et al. 2004; Varoqueaux, Jamain et al. 2004; Levinson, Chery et al. 2005).  Knock-down of these proteins  results in a reduction of both excitatory and inhibitory synapses (Chih, Engelman et al. 2005). It remains unclear how NL retention/function is modulated at inhibitory sites. One possibility is that molecules such as gephyrin may act to regulate NL accumulation at inhibitory contacts.  If the preformed scaffold participates in excitatory synapse development, then disruption of this complex may result in a decrease in the number of excitatory synapses. Indeed, loss of PSD-95 results in a reduction of GKAP and Shank clusters as well as a decrease in VGLUT positive sites indicating a decrease in excitatory contact number.  The reduction of GKAP, Shank and VGLUT clusters correlated with an increase in number of VGAT sites containing NL1. This may explain our previous observation that knock down of PSD-95 decreases the excitatory to inhibitory synaptic ratio (Prange et al. 2004).  PSD- 95 knock down only results in partial loss of excitatory synapses and partial redistribution of some of the associated proteins, thus other factors must be involved for excitatory synapse development. In contrast, animals mutant for PSD-95 exhibit normal basal excitatory transmission, indicating that PSD-95 is not essential for excitatory synapse formation, but recent data by Sheng and colleagues showed that acute knock down of either PSD-95 or SAP-97 diminishes excitatory synapse transmission and glutamate receptor clustering at the synapse (Nakagawa et al. 2004; Prange et al. 2004).  Thus, future studies are required to clarify whether functional redundancy has compensated for the loss of PSD-95 in mutant animals or whether these proteins serve functions unrelated to synapse formation. Despite this controversy, data presented here reveal that a preformed non-synaptic complex of these proteins can precede presynaptic maturation.  We propose a mechanism by which stationary clusters of postsynaptic proteins may serve to regulate the number and location of synapses formed at early stages of synaptogenesis (Figure 2.16). A pre-assembled protein complex may help guiding axons to form functional presynaptic contacts. The constituents of this protein complex will also help determine whether the nascent contact is an excitatory or inhibitory synapse. The  102 observation that transport packets of the presynaptic protein synaptophysin can be delivered to these sites, suggest that the preassembled complex of postsynaptic proteins may instruct delivery of presynaptic components required for synthesis, transport and release of glutamate. The number of these stationary sites may then determine the location and number of new excitatory synapses formed. Such a mechanism may be required for the precise assembly of postsynaptic elements that perfectly match the identity of the neurotransmitter to be recruited to presynaptic terminals. In the future it will be important to determine how neurons control the number and location for docking the preassembled protein complex.   Figure 2.16: A model summarizes the role of stationary and mobile preformed scaffold clusters in excitatory synapse development. Stationary (orange) and mobile (yellow) preformed scaffolding clusters are abundant early in development, with a few mature postsynaptic densities (brown) apposed to functional presynaptic terminals. Stationary preformed scaffold clusters are NLG1 positive, and can attract immature presynaptic terminals. NLG1 can drive recruitment of synaptophysin transport packets. Mobile preformed clusters lack NLG1 and are transported to stationary preformed scaffolding clusters or mature postsynaptic sites. Splitting and mobile clusters can act as a reservoir that generates new stationary clusters. These in turn serve as new hotspots for development of excitatory synapses. This model postulates a postsynaptic mechanism whereby preassembled non-synaptic clusters containing scaffolding proteins and cell adhesion molecules dictate the location and number of newly formed excitatory contacts.  103  2.5 REFERENCES  Ahmari, S. E., and Smith, S. J. (2002). Knowing a nascent synapse when you see it. Neuron 34, 333-336.  Biederer, T., Sara, Y., Mozhayeva, M., Atasoy, D., Liu, X., Kavalali, E. T., and Sudhof, T. C. (2002). SynCAM, a synaptic adhesion molecule that drives synapse assembly. Science 297, 1525-1531.  Blue, M. E., and Parnavelas, J. G. (1983). The formation and maturation of synapses in the visual cortex of the rat. I. Qualitative analysis. J Neurocytol 12, 599-616.  Bredt, D. S., and Nicoll, R. A. (2003). AMPA receptor trafficking at excitatory synapses. Neuron 40, 361- 379.  Bresler, T., Ramati, Y., Zamorano, P. L., Zhai, R., Garner, C. C., and Ziv, N. E. (2001). The dynamics of SAP90/PSD-95 recruitment to new synaptic junctions. Mol Cell Neurosci 18, 149-167.  Bresler, T., Shapira, M., Boeckers, T., Dresbach, T., Futter, M., Garner, C. C., Rosenblum, K., Gundelfinger, E. D., and Ziv, N. E. (2004). Postsynaptic density assembly is fundamentally different from presynaptic active zone assembly. J Neurosci 24, 1507-1520.  Chih, B., Engelman, H., and Scheiffele, P. (2005). Control of Excitatory and Inhibitory Synapse Formation by Neuroligins. Science 307, 1324-1328.  Craig, A. M., and Boudin, H. (2001). Molecular heterogeneity of central synapses: afferent and target regulation. Nat Neurosci 4, 569-578.  El-Husseini, A. E., Craven, S. E., Chetkovich, D. M., Firestein, B. L., Schnell, E., Aoki, C., and Bredt, D. S. (2000a). Dual palmitoylation of PSD-95 mediates its vesiculotubular sorting, postsynaptic targeting, and ion channel clustering. J Cell Biol 148, 159-172.  El-Husseini, A. E., Schnell, E., Chetkovich, D. M., Nicoll, R. A., and Bredt, D. S. (2000b). PSD-95 involvement in maturation of excitatory synapses. Science 290, 1364-1368.  Fiala, J. C., Feinberg, M., Popov, V., and Harris, K. M. (1998). Synaptogenesis via dendritic filopodia in developing hippocampal area CA1. J Neurosci 18, 8900-8911.  Friedman, H. V., Bresler, T., Garner, C. C., and Ziv, N. E. (2000). Assembly of new individual excitatory synapses: time course and temporal order of synaptic molecule recruitment. Neuron 27, 57-69.  Fu, Z., Washbourne, P., Ortinski, P., and Vicini, S. (2003). Functional excitatory synapses in HEK293 cells expressing neuroligin and glutamate receptors. J Neurophysiol 90, 3950-3957.  Garner, C. C., Zhai, R. G., Gundelfinger, E. D., and Ziv, N. E. (2002). Molecular mechanisms of CNS synaptogenesis. Trends Neurosci 25, 243-251.  Graf, E. R., Zhang, X., Jin, S. X., Linhoff, M. W., and Craig, A. M. (2004). Neurexins Induce Differentiation of GABA and Glutamate Postsynaptic Specializations via Neuroligins. Cell 119, 1013- 1026.  Hinds, J. W., and Hinds, P. L. (1976). Synapse formation in the mouse olfactory bulb. II. Morphogenesis. J Comp Neurol 169, 41-61.   104 Kennedy, M. B. (1997). The postsynaptic density at glutamatergic synapses. Trends Neurosci 20, 264-268.  Kim, E., and Sheng, M. (2004). PDZ domain proteins of synapses. Nat Rev Neurosci 5, 771-781.  Krueger, S. R., Kolar, A., and Fitzsimonds, R. M. (2003). The presynaptic release apparatus is functional in the absence of dendritic contact and highly mobile within isolated axons. Neuron 40, 945-957.  Levinson, J. N., Chery, N., Huang, K., Wong, T. P., Gerrow, K., Kang, R., Prange, O., Wang, Y. T., and El-Husseini, A. (2005). Neuroligins Mediate Excitatory and Inhibitory Synapse Formation: INVOLVEMENT OF PSD-95 AND NEUREXIN-1{beta} IN NEUROLIGIN-INDUCED SYNAPTIC SPECIFICITY. J Biol Chem 280, 17312-17319.  Li, Z., and Sheng, M. (2003). Some assembly required: the development of neuronal synapses. Nat Rev Mol Cell Biol 4, 833-841.  Marrs, G. S., Green, S. H., and Dailey, M. E. (2001). Rapid formation and remodeling of postsynaptic densities in developing dendrites. Nat Neurosci 4, 1006-1013.  Migaud, M., Charlesworth, P., Dempster, M., Webster, L. C., Watabe, A. M., Makhinson, M., He, Y., Ramsay, M. F., Morris, R. G., Morrison, J. H., et al. (1998). Enhanced long-term potentiation and impaired learning in mice with mutant postsynaptic density-95 protein. Nature 396, 433-439.  Naisbitt, S., Valtschanoff, J., Allison, D. W., Sala, C., Kim, E., Craig, A. M., Weinberg, R. J., and Sheng, M. (2000). Interaction of the postsynaptic density-95/guanylate kinase domain-associated protein complex with a light chain of myosin-V and dynein. J Neurosci 20, 4524-4534.  Nakagawa, T., Futai, K., Lashuel, H. A., Lo, I., Okamoto, K., Walz, T., Hayashi, Y., and Sheng, M. (2004). Quaternary structure, protein dynamics, and synaptic function of SAP97 controlled by L27 domain interactions. Neuron 44, 453-467.  Okabe, S., Miwa, A., and Okado, H. (2001). Spine formation and correlated assembly of presynaptic and postsynaptic molecules. J Neurosci 21, 6105-6114.  Passafaro, M., Nakagawa, T., Sala, C., and Sheng, M. (2003). Induction of dendritic spines by an extracellular domain of AMPA receptor subunit GluR2. Nature 424, 677-681.  Prange, O., and Murphy, T. H. (1999). Correlation of miniature synaptic activity and evoked release probability in cultures of cortical neurons. J Neurosci 19, 6427-6438.  Prange, O., and Murphy, T. H. (2001). Modular transport of postsynaptic density-95 clusters and association with stable spine precursors during early development of cortical neurons. J Neurosci 21, 9325- 9333.  Prange, O., Wong, T. P., Gerrow, K., Wang, Y. T., and El-Husseini, A. (2004). A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. PNAS 101, 13915-13920.  Rao, A., Kim, E., Sheng, M., and Craig, A. M. (1998). Heterogeneity in the molecular composition of excitatory postsynaptic sites during development of hippocampal neurons in culture. J Neurosci 18, 1217- 1229.  Romorini, S., Piccoli, G., Jiang, M., Grossano, P., Tonna, N., Passafaro, M., Zhang, M., and Sala, C. (2004). A functional role of postsynaptic density-95-guanylate kinase-associated protein complex in regulating Shank assembly and stability to synapses. J Neurosci 24, 9391-9404.  Sabo, S. L., and M cAllister, A. K. (2003). Mobility and cycling of synaptic protein-containing vesicles in axonal growth cone filopodia. Nat Neurosci 6, 1264-1269.  105 Sala, C., Piech, V., Wilson, N. R., Passafaro, M., Liu, G., and Sheng, M. (2001). Regulation of dendritic spine morphology and synaptic function by Shank and Homer. Neuron 31, 115-130.  Sanes, J. R., and Lichtman, J. W. (2001). Induction, assembly, maturation and maintenance of a postsynaptic apparatus. Nat Rev Neurosci 2, 791-805.  Sans, N., Prybylowski, K., Petralia, R. S., Chang, K., Wang, Y. X., Racca, C., Vicini, S., and Wenthold, R. J. (2003). NMDA receptor trafficking through an interaction between PDZ proteins and the exocyst complex. Nat Cell Biol 5, 520-530.  Scheiffele, P., Fan, J., Choih, J., Fetter, R., and Serafini, T. (2000). Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell 101, 657-669.  Shapira, M., Zhai, R. G., Dresbach, T., Bresler, T., Torres, V. I., Gundelfinger, E. D., Ziv, N. E., and Garner, C. C. (2003). Unitary assembly of presynaptic active zones from Piccolo-Bassoon transport vesicles. Neuron 38, 237-252.  Steward, O., and Falk, P. (1986). Protein-synthetic machinery at postsynaptic sites during synaptogenesis: a quantitative study of the association between polyribosomes and developing synapses. J Neurosci 6, 412- 423.  Tu, J. C., Xiao, B., Naisbitt, S., Yuan, J. P., Petralia, R. S., Brakeman, P., Doan, A., Aakalu, V. K., Lanahan, A. A., Sheng, M., and Worley, P. F. (1999). Coupling of mGluR/Homer and PSD-95 complexes by the Shank family of postsynaptic density proteins. Neuron 23, 583-592.  Varoqueaux, F., Jamain, S., and Brose, N. (2004). Neuroligin 2 is exclusively localized to inhibitory synapses. Eur J Cell Biol 83, 449-456.  Waites, C. L., Craig, A. M., and Garner, C. C. (2005). Mechanisms of vertebrate synaptogenesis. Annu Rev Neurosci 28, 251-274.  Washbourne, P., Bennett, J. E., and McAllister, A. K. (2002). Rapid recruitment of NMDA receptor transport packets to nascent synapses. Nat Neurosci 5, 751-759.  Washbourne, P., Liu, X. B., Jones, E. G., and McAllister, A. K. (2004). Cycling of NMDA receptors during trafficking in neurons before synapse formation. J Neurosci 24, 8253-8264.  Ziv, N. E., and Garner, C. C. (2004). Cellular and molecular mechanisms of presynaptic assembly. Nat Rev Neurosci 5, 385-399.         106  3 Neuroligin trafficking on the surface: The role of diffusion in the trafficking of adhesion proteins.                                 A version of this chapter has been submitted for publication. Gerrow K*, Takahashi H*, Howarth M, Ting AY, El-Husseini A. (2008) Neuroligin trafficking on the surface: The role of diffusion in the trafficking of adhesion proteins. *these authors contributed equally.  107 3.1 INTRODUCTION  From a molecular perspective, synapse formation involves the assembly of hundreds of pre- and postsynaptic proteins in order to function properly. The adhesion complex of neuroligin (NL) and neurexin (NRX) can trigger assembly of glutamatergic (excitatory) and GABAergic (inhibitory) proteins in cultured hippocampal neurons (Chih et al., 2005; Graf et al., 2004; Levinson et al., 2005; Prange et al., 2004; Sara et al., 2005; Scheiffele et al., 2000). Immuno-localization studies demonstrate that different neuroligin isoforms are preferentially targeted: NL1 is primarily localized to the postsynaptic side of glutamatergic synapses, whereas NL2 is preferentially targeted to GABAergic synapses (Song et al., 1999; Varoqueaux et al., 2004; Levinson et al., 2005; Chih et al., 2005). The mode of trafficking and retention of adhesion proteins to synapses remain unclear, and mechanisms that influence targeting to excitatory versus inhibitory are not fully understood. Therefore we used NL1 and NL2, two adhesion proteins with high sequence homology and differential targeting to glutamatergic and GABAergic synapses respectively, to explore the hypothesis that adhesion molecules accumulate at synapses through lateral diffusion on the membrane surface. Furthermore we predict that molecular mechanisms which influence this surface mobility may help determine synaptic targeting. Synapse formation and maturation in the CNS has been extensively studied using a model system of glutamatergic synapses in hippocampal or cortical neurons (reviewed in McAllister, 2007; Waites et al., 2005; Ziv and Garner, 2004)). Formation of presynaptic active zones occurs by insertion of precursor vesicles containing µpre-packaged¶ active zone components, such as bassoon and piccolo, and subsequent recruitment of synaptic vesicles (Bresler et al., 2004; Sabo et al., 2006; Shapira et al., 2003). Assembly of the postsynaptic apparatus can occur via sequential recruitment of molecules to the postsynaptic membrane and their assimilation in situ (Bresler et al., 2001; Bresler et al., 2004; Okabe et al., 1999), but can also occur by the assembly of preformed complexes of scaffolding proteins (Gerrow et al., 2006) and/or receptors (Washbourne et al., 2002). The formation and co-ordination of glutamatergic synapses can be influenced by several classes of proteins: Diffusible factors such as Narp family of proteins; signaling cascades  108 such as those BDNF-TrK; adhesion complexes such as neuroligin/neurexins and SynCAM; and scaffolding proteins such as PSD-95 and other members of the MAGUK family. These proteins can work in a hierarchical fashion, or in parallel in order to mediate synapse formation and maturation. The assembly of GABAergic synapses is less understood. Previous studies on GABAergic synapse formation have focused on the clustering of GABAA receptors and gephyrin, a putative postsynaptic scaffold protein (Dumoulin et al., 2000; Moss and Smart, 2001 Christie et al., 2002b; Danglot et al., 2003; Luscher and Keller, 2004; Studler et al., 2005). GABAergic synapses are detected earlier than glutamatergic synapses during embryonic brain development and in dissociated neuron cultures (Tyzio et al., 1999; Khazipov et al., 2001; Hennou et al., 2002; Deng et al., 2007). GABAA receptors form clusters on cell membranes even before synapse formation (Brunig et al., 2002; Christie et al., 2002a; Danglot et al., 2003; Studler et al., 2005), and are often smaller than those induced by GABAergic nerve terminals (Brunig et al., 2002; Christie et al., 2002a). Gephyrin is a cytoplasmic protein that accumulates at the postsynaptic complex of GABAergic and glycinergic synapses where it forms sub-membranous lattices associated with postsynaptic clusters of receptors (Kneussel and Betz, 2000). Studies with a gephyrin-deficient mice have shown that while gephyrin is essential for the synaptic clustering of glycine receptors (Essrich et al., 1998; Feng et al., 1998; Levi et al., 2004), gephyrin is only essential for the clustering of some GABAARs (Kneussel et al., 2001; Kneussel et al., 1999; Levi et al., 2004). In cultured hippocampal pyramidal cells, knock- down of gephyrin decreased the number of GABA(A) receptor clusters at postsynaptic sites (Yu et al., 2007). In contrast to receptors, the role and mode of adhesion complex trafficking to GABAergic synapses remains enigmatic. For both glutamatergic and GABAergic synapses, the number of neurotransmitter receptors at synapses likely depends on both the regulation of diffusion in the plasma membrane and the capacity of synapses to capture receptors. We hypothesize that adhesion molecules, and in particular NLs, may also be trafficked in a similar manner. In order to test this hypothesis, we performed fluorescence recovery after photo-beaching (FRAP), where the rate of recovery of fluorescence of a bleached spot is then monitored  109 over time. Under this paradigm, fluorescence recovery occurs through mobility of proteins from the area surrounding the bleached spot. We show that FRAP dynamics of NL1 and NL2 are significantly different at sites which contain their 'synapse-appropriate' scaffold. Interestingly, FRAP dynamics of AP-tagged NL clusters labeled by BirA, a method that specifically labels surface proteins, are similar to GFP-tagged proteins, suggesting that most mobile NL comes from a surface pool. The mobility of NL on the surface was influenced by intracellular domains and interactions with specific scaffolding proteins and the cytoskeleton. For instance, the exogenous expression of certain scaffolding proteins was able to decrease the mobile pool, thus effectively trapping NLs at their appropriate site. Together these results demonstrate the lateral trafficking of NLs associated with excitatory and inhibitory scaffold sites, and dissect several of the molecular interactions that influence this trafficking within the membrane.    110 3.2 MATERIALS AND METHODS  3.2.1 cDNA constructs  PSD-95 GFP was constructed by PCR and sub-cloned in frame into HindIII and EcoRI restriction sites of GW1-GFP. Shank-YFP were generated as previously described (Romorini et al., 2004). GFP S-SCAM was a gift from Dr. E Kim and was made as described (Iida et al., 2004). Neuroligin-1 CFP was a gift from Dr. P. Washbourne (University of Oregon) and was made as described (Fu et al., 2003). Neuroligin-2 GFP was constructed by PCR and sub-cloned in frame into eGFP-C1.  AP-tagged neuroligin-1 (AP NL1) and AP-tagged neuroligin -2, (AP NL2) was constructed from HA NL1 (Levinson et al., 2005), where HA was excised with NheI And AgeII, and replaced with an AP tag. AP-NL2 mutants were similarly generated from HA- tagged versions that were previously generated by J. Levinson. Construction of NL2CT was carried out by PCR of full-length NL2 using oligonucleotides with BglII and HinDIII restriction sites (forward: GGGCCCAGATCTCGGGGAGGAGGGGGTCCC; reverse: GGGCCCAAGCTTCTATGGGCTAAGCCGCCTGCACCGCAG), and subsequently sub-cloning the resultant fragment into full-length HA-NL2 with the corresponding BglII/HinDIII fragment removed. This resulted in an HA-tagged construct encoding amino acids 1-715 of NL2. NL2∆PDZb and NL2∆PR were constructed in the same manner, with the exception that the following reverse primers were used: GGGCCCAAGCTTCTAGGAGTGGGGATGGGG for NL2PDZb; and GAGTCCAAGCTTCTAT ACCCGAGTGGTGGAGTGGGGATGGGGTAGCCCCAGGCCACTGGGCAGCAGGGTC for NL2PR. This generated an HA-tagged NL2 construct either lacking its C-terminal four amino acids, which encode the PDZ-binding domain, or lacking the region from amino acids 798-826, which encodes the proline-rich region. NL2 lacking the membrane- proximal region (∆MR, 714-782) was constructed in two steps. First, PCR of full-length  111 NL2 was conducted using the following primers: forward primer: GGACTCAAGCTTCTCTTGGCCCCGGGGGCC; reverse primer: GGGCCCGTCGACCTATACCC GAGTGGTGGAGTGGGGATGGGGTAG. The resulting fragment, corresponding to amino acids 783-836 of NL2, was then subcloned into HA-NL2∆CT, using HinDIII and SalI restriction sites. Next, the stop codon remaining from NL2∆CT was removed using the Quick-change site-directed mutagenesis kit (Stratagene).  Briefly, the entire plasmid was amplified by PCR using oligonucleotides containing a deletion for TAG (GGCGGCTTAGCCCAAAGCTTCTCTTGGC), the methylated template strand was destroyed enzymatically with DpnI, and the purified construct transformed into bacteria. All constructs have been verified by sequencing.   3.2.2 Neuronal cell culture and transfection  Dissociated primary neuronal cultures were prepared from hippocampi of embryonic day 18 Wistar rats and maintained in NeuroBasal media (NBM, Invitrogen) supplemented with B-27, penicillin, streptomycin, and L-glutamine. Neurons were plated at 80,000 per 12 mm glass cover slip, or 500,000 per 35mm culture dish (Corning, Matek). Transfection was preformed on DIV 5-8 using 1.6µg DNA and 2µl of Lipofectamine 2000, and imaged 20-36 hours later for near exogenous levels. Under these conditions, protein over-expression was minimal, did not enhance presynaptic terminals, and localized correctly (Appendix 5.3). For mutants and co-transfection with scaffold proteins, expression was continued for 2 days in order to distinguish changes in trafficking.       112 3.2.3 Immunocytochemistry and analysis  Coverslips were fixed in ±20C methanol and immuno-labeled for synaptic proteins. Mouse monoclonal antibodies include: neuroligin (1:500, Synaptic Systems). Rabbit polyclonal antibodies include: synaptophysin (1:1000; PharMingen), VGLUT1 (1:1000; Synaptic Systems). Chicken polyclonal antibodies include: (1:1000; Chemicon), and GFP (1:2000). Secondary antibodies were generated in goat, and conjugated with Alexa488 (1:1000), Alexa568 (1:1000), or Alexa360 (1:400; Molecular Probes) All antibody reactions were preformed in blocking solution (2% normal goat serum) for 1 hour at room temperature or overnight at 4 0C. Images were scaled to 16 bits and analyzed in Northern Eclipse (Empix Imaging, Missasauga, Canada), by using user-written software routines (Appendix 5.2). Two-tailed parametric Student¶s T-test was preformed to calculate statistical significance of results between experimental groups.   3.2.4 Surface labeling with monovalent streptavidin  Following enzymatic biotinylation of AP-tagged neuroligin in the endoplasmic reticulum by bacterially derived biotin ligase (BirA) with an ER retention site (KDEL), surface labeling with monovalent streptavidin (see Appendix 5.5 for synthesis) was preformed as previously described (Chen et al. 2005, Howarth et al. 2005, 2006). Briefly, neurons were incubated for 5 min with 5 µg/ml Alexa Fluor 568±conjugated monovalent streptavidin in Hank¶s Balanced Salt Solution (HBS, Invitrogen), washed with supplemented NBM, and used immediately for live imaging analysis. The use of a monovalent streptavidin allows for stable and cross-linking±free protein detection on the cell surface in imaging experiments. (Figure 3.1, Table 3.1). In addition, our experiments with neuroligin-1 demonstrate how engineered streptavidins can be used to examine the biological consequences of controlled clustering of cellular proteins (Figure 3.2, Table 3.2).    113   Figure 3.1: Monovalent streptavidin reduces aggregation of NL1. Hippocampal neurons in dissociated culture were transfected with AP±neuroligin-1, biotinylated with biotin ligase and labeled with Alexa Fluor 568±conjugated wild-type (left) or A1D3 (right) streptavidin. After staining, cells were incubated for 0 h (top) or 2 h (middle) at 37 ƒC and streptavidin staining was visualized by live-cell fluorescence microscopy. Scale bar, 10 m. Magnified images of the boxed regions are shown at the bottom. Scale bar, 1 m.    Table 3.1: NL1 clustering by streptavidin. Labeling condition n Neuroligin clusters per 10 µm dendrite Significance A1D3 (2 h, live cells) 8 0.9 ± 0.2 Wild type (2 h, live cells) 10 2.2 ± 0.3 p = 0.002 A1D3 (2 h, fixed cells) 9 0.9 ± 0.2 Wild type (2 h, fixed cells) 10 2.7 ± 0.2 p = 5x10-6 Hippocampal neurons expressing AP-neuroligin-1 were biotinylated, labeled with A1D3 or wild type streptavidin, and incubated for 2 hours. Neurons were then imaged live or after fixation and the number of AP-neuroligin-1 clusters was determined (± s.e.m.). Wild type streptavidin and A1D3 samples were prepared and imaged under identical conditions.    114  Figure 3.2: Aggregation of NL1 by wild-type streptavidin reduces NL1 enhancement of presynaptic terminals. Neurons were biotinylated and labeled with wild-type or A1D3 streptavidin as in a, incubated for 24 h and then stained for the presynaptic marker VGLUT1. Streptavidin (red) and VGLUT1 (green) signals are shown separately or overlaid. Scale bar, 10 m. Magnified images of the boxed regions are shown below. Scale bar, 1 m. Arrows indicate AP±neuroligin-1 clusters not apposed to presynaptic termini.   Table 3.2: Effect of streptavidin on VGLUT1 clustering. Labeling condition n VGLUT1 clusters  per 10 µm dendrite Fold-enhancement of VGLUT1 intensity A1D3 (24 h) 20 5.8 ± 0.4 2.3 ± 0.1 Wild type (24 h) 10 4.3 ± 0.4 p = 0.05 1.8 ± 0.1 p = 0.0002 HA-neuroligin-1  11 6.2 ± 0.6  2.2 ± 0.1 Wild type (0 h) 11 5.7 ± 0.4  2.2 ± 0.3 Hippocampal neurons expressing AP-neuroligin-1 were biotinylated and A1D3 or wild type streptavidin was added for 24 hours or just prior to fixation (0 h). Neurons were fixed and the number and intensity of VGLUT1 clusters were quantified (± s.e.m.). The intensity of VGLUT1 clusters is given relative to untransfected neurons in the same field of view. The analysis of VGLUT1 clusters from neurons transfected with HA-neuroligin-1 is also shown. Wild type streptavidin and A1D3 samples were prepared and imaged under identical conditions.  115 3.2.5 Fluorescence recovery after photo-beaching (FRAP) All FRAP images were acquired in an environmentally controlled chamber (37ƒC, 5% CO2) on a Zeiss LSM 510 META with a 63x Achroplan (0.95NA) ceramic dipping lens. After the collection of baseline images, photo-bleaching was performed by scanning two pre-selected 18 x 18 pixel (approximately 2.5 x 2.5 µm) regions of interest sequentially at the appropriate wavelength (458nm for CFP, 488 nm for GFP, and 543 nm for Alexa568), at high illumination intensity for 80-120 iterations. After the photo-bleaching procedure, fluorescence recovery was recorded by automated time-lapse microscopy, with images being acquired every 20-70 seconds. Only clusters in which fluorescence was reduced to at least 50% of original fluorescence, and showed less than 15% of photo- bleaching due to time-lapse imaging were used in the final analysis. Data was collected from 2-4 independent experiments, 8-15 neurons, and 12-30 clusters per group unless otherwise indicated. Mean fluorescence intensity was measured for each cluster using Image J, and off-cell background intensity was subtracted. An area was also measured on-cell for adjustments in bleaching and laser fluctuations. To combine experiments, pre- bleach values were set to 100%, and immediate post-bleach intensities to 0%. The mean curve of FRAP for each cluster was fitted by GraphPad Prism software using a non-linear regression equation with one phase exponential association to determine Ymax (mobile fraction), K (rate constant) and t1/2. Statistical analysis was performed using a non- parametric student¶s t-test where n represents the number of clusters analyzed. 3.2.6 Time-lapse Imaging was preformed 24-36 hours post-transfection in an environmentally controlled stage (37 0C and 5% CO2). Images were acquired on a Zeiss Axiovert M200 motorized microscope with a 63X 1.4 NA ACROMAT oil immersion lens and a monochrome 14-bit Zeiss Axiocam HR charged-coupled camera with 1300 X 1030 pixels.  Exposures were preformed at 1/3 saturation (200-800ms) to minimize photo-damage to live cells. To correct for out-of-focus clusters within the field of view, focal plane (z-) stacks were acquired and maximum intensity projections performed offline.  116 3.3 RESULTS  3.3.1 Neuroligin mobility is isoform specific and developmentally regulated  The relationships between the constant flux of the constitutive elements, and the structural stability of synapses are still not understood. We hypothesize that the trafficking of NL adhesion proteins to the appropriate site involves lateral diffusion from extra-synaptic sites to synaptic sites. If true, dynamics of NL into and out of these sites can be determined by fluorescence recovery after photo-beaching (FRAP). The amount of NL able to exchange would be represented by the existence of a mobile fraction, and the kinetics of this exchange would be represented by a time constant. Thus, a stable synaptic component would have a small mobile pool with a large time constant, and vice versa for a dynamic synaptic component. To test for the lateral diffusion of NL, experiments were performed on fluorescently-tagged NL1 at Shank sites, and NL2 at gephyrin sites using conditions that minimize over-expression (Appendix Figure 5.5), and ensure targeting similar to endogenous proteins (Appendix  Figure 5.6) (Gerrow et al., 2005; Kuriu et al., 2006; Washbourne et al., 2004). Importantly, co-expression of fluorescently-tagged marker scaffolds (Shank and gephyrin), did not significantly change the mobile fraction of NL clusters compared to GFP co-transfected neurons (Appendix Figure 5.7).  At day in vitro 7 (DIV7), a developmental stage that corresponds with the beginning of synapse formation and when most synapses are on dendritic shafts, recovery of NL1 and NL2 (Figure 3.3 A,B) had similar rates of recovery (t 1/2 133.3±22.4s and 220.2±47.2s respectively, p=0.07). Interestingly, the mobile pool of NL1 (74.5±7.8%) was significantly greater than NL2 (49.0±5.5%, p<0.001, Figure 3.3 C), suggesting that NL1 at Shank sites is more dynamic, and potentially more malleable, than NL2 at gephyrin sites.    117  Figure 3.3: Neuroligin mobility is isoform specific. Fluorescence recovery after photo-beaching (FRAP) on fluorescently-tagged NL1 at Shank sites, and NL2 at gephyrin sites after 1 day of expression. (A) Representative example of a site where CFP-NL1 and YFP- Shank1b were co-clustered at DIV 7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Site where GFP-NL2 and CFP-gephyrin were co-clustered (left panels) were bleached and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of NL1 and NL2 at sites containing their appropriate scaffold. Pre-bleach values were normalized to 100, and post-bleach values (t=0) were normalized to 0 in order to compare groups. The mobile fraction of NL1 at Shank sites was 74.5±7.8% (n=20) was significantly greater than NL2 at gephyrin sites 49.0±5.5% (n=16), p<0.001. Scale bars, 1µm. Error bars indicate S.E.M .   At DIV14, a developmental stage that corresponds with the rapid synapse formation and maturation, the mobile fraction of NL1 was slightly less than compared to DIV 7 (Figure 3.4 A,B; mobile pool 54.9±4.4%,p<0.05). NL2 recovery was also blunted, as indicated by a significant reduction in the mobile pool (Figure 3.4 C,D; 27.8±9.7%, p<0.05). These results indicate that the mobile fraction of NL1 and NL2 are significantly different at sites with their synapse appropriate scaffold, and that this mobile fraction is reduced at later developmental stages.   118  Figure 3.4: Neuroligin mobility is developmentally regulated. (A) Representative example of a site where CFP-NL1 and YFP-Shank1b were co-clustered at DIV 14 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Site where GFP-NL2 and CFP-gephyrin were co-clustered at DIV 12 (left panels) were bleached and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of NL1 at DIV7 versus DIV14. The mobile fraction of NL1 at Shank sites at DIV14 was 54.9±4.4% (n=20), and significantly less than at DIV7 (p<0.05). (D) FRAP curves comparing the recovery of NL2 at DIV7 versus DIV12. The mobile fraction of NL2 at gephyrin sites at DIV12 was 27.8±9.7% (n=18), and significantly less than at DIV7 (p<0.05). Scale bars, 1µm. Error bars indicate S.E.M .   3.3.2 Surface neuroligin accounts for most of the mobile pool.  FRAP was initially designed to measure two-dimensional diffusion of membrane-bound molecules (Axelrod et al., 1976), and indeed, the rapid nature of the recovery after bleaching of fluorescently-tagged neuroligins suggests the existence of a mobile pool at the surface. However, the use of fluorescently-tagged proteins does not rule out the possibility that recovery could be due to insertion of protein from an intracellular source. In order to distinguish this, we have taken advantage of a site-specific biotinylation-based approach that allows for visualization of surface proteins in live neurons with monovalent streptavidin (mSA). We generated versions of NL1 and NL2 containing a 15 amino acid acceptor peptide (AP), which were co-expressed with a bacterially derived biotin ligase (BirA) containing the endoplasmic reticulum (ER) retention motif KDEL. Under these conditions, BirA uses ATP and biotin from the neuron to specifically catalyze the  119 addition of biotin onto the AP-sequence in the ER. Surface neuroligins can then be visualized by incubation with fluorescently conjugated monovalent streptavidin (Alexa- mSA), an important reagent for eliminating cross-linking artifact (Howarth et al., 2006). Using cell surface site-specific biotinylation of NL proteins, we saw a 66.2±3.7% recovery of AP-tagged NL1 at Shank sites and 36.2±4.3% recovery after bleaching of AP-tagged NL2 at gephyrin sites (Figure 3.5 A,B). The dynamics of FRAP of AP-tagged neuroligins was similar to their fluorescently-tagged counterparts (Figure 3.5 C,D; p=0.3 and 0.1 for NL1 and NL2 respectively), suggesting that the mobile pool of neuroligins is made of surface proteins.    Figure 3.5: Surface neuroligin accounts for most of the mobile pool. FRAP of AP-tagged NL1 at Shank sites, and NL2 at gephyrin sites after 1 day of expression. (A) Representative example of a site where AP-NL1 and YFP-Shank1b were co-clustered at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Site where AP-NL2 and CFP-gephyrin were co-clustered (left panels) were bleached and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of CFP-NL1 and AP-NL1 at DIV7. The mobile fraction of APNL1 was 66.2±3.7% (n=18), and not significantly different from CFP-NL1 (p=0.3). Scale bars, 1µm. Error bars indicate S.E.M .   The age dependence of NL recovery was maintained when looking at just the surface pool (Figure 3.6 A,B). At DIV 12, the recovery of NL1 was significantly reduced (Figure 3.6 C,E; mobile fraction 38.5±4.3%) compared to DIV 7 (p<0.001), as well as the  120 recovery of NL2 (Figure 3.6 D; mobile fraction 32.2±2.9%, p<0.001). Since FRAP dynamics of AP-tagged NL clusters labeled by BirA, a method that specifically labels surface proteins, are similar to GFP-tagged proteins, these data suggest that most mobile neuroligin comes from a surface pool, and that there is a reduction in the mobile fraction at the surface during development. Thus, mechanisms that decrease the mobile pool of NLs, leading to a more sustained retention at specific sites, may influence the maturation of synapses.    Figure 3.6: Neuroligin mobility on the surface is developmentally regulated. (A) Representative example of a site where AP NL1 and YFP-Shank1b were co-clustered at DIV12 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Site where APNL2 and CFP-gephyrin were co-clustered at DIV12 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of GFP-NL2 and AP-NL2 at DIV7. The mobile fraction of APNL1 was 36.2±4.3% (n=16), and not significantly different from CFPNL1 (p=0.1). (D) FRAP curves comparing the recovery of APNL1 at DIV7 versus DIV12 after 2 days expression. The mobile fraction of APNL1 at Shank sites at DIV12 was 38.5±4.3% (n=18), and significantly less than at DIV7 (p<0.001). (F) FRAP curves comparing the recovery of NL2 at DIV7 versus DIV12 after 2 days expression. The mobile fraction of NL2 at gephyrin sites at DIV12 was 32.2±2.9% (n=23), and significantly less than at DIV7 (p<0.05). Scale bars, 1µm. Error bars indicate S.E.M.   121 3.3.3 Recovery of scaffolding proteins and neuroligins are correlated. Next we wondered whether FRAP of scaffold proteins paralleled that of NLs. Indeed, the size of the mobile fraction of Shank at clusters (75.2±3.2%) was similar to NL1, and the size of the mobile fraction of gephyrin at clusters (29.1±3.8%) was similar to NL2 (Figure 3.7 A,B). When compared to each other at the same sites, the mobile fraction and time constants of Shank and NL1 have a high degree of correlation (Figure 3.7 C,D; Pearson¶s r=0.56 and 0.72, p<0.01 and p<0.001 respectively). This correlation between NL1 and Shank suggests that the amount and rate of dissociation/association of NL1 and Shank may be related, sharing a similar underlying mechanism for their association at these sites. This correlation was also true for the mobile fraction of gephyrin and NL2, although not as strong and only apparent when comparing mobile fractions (r=0.43, p<0.05). These data suggest that mechanisms determining the size of the mobile pool of NL1 at these sites may similarly influence Shank.   Figure 3.7: Correlation of NL recovery at the correct scaffold. (A) FRAP of AP-tagged NL1 and Shank at the same site (upper panels), and AP NL2 and gephyrin at the same site (lower panels). (B) FRAP curves comparing the recovery of Shank and gephyrin at DIV7. The mobile fraction of Shank is 75.2±3.2% (n=30), and the mobile fraction of gephyrin is 29.1±3.8% (n=30). (C) Correlation of the mobile fraction (r=0.56, p<0.01) and time constant (r=0.72, p<0.001) for NL1 and Shank co-clusters. (D) Correlation of the mobile fraction (r=0.43, p<0.05) and time constant (r=0.29) for NL2 and gephyrin co-clusters. Scale bars ѽҏңµm. Error bars indicate S.E.M .   122  3.3.4 Neuroligin recovery at sites where scaffold proteins are diffuse.  Previous experiments with neurotransmitter receptors from both glutamatergic and GABAergic synapses, have shown that the mobile fraction of these proteins is significantly larger at extra-synaptic sites (Jacob et al., 2005). It has been suggested that this larger mobile fraction is due to a lack of interactions to anchor and decrease the mobility of receptors. If similar mechanisms influence mobility of NL, then similarly, a larger mobile fraction should be apparent at extra-synaptic sites where scaffolding proteins are not clustered. In order to determine if the mobile pool of neuroligins when diffuse and not associated with the appropriate scaffold we performed FRAP of AP- tagged NL1 and NL2 not associated with scaffold proteins and in parallel, at clustered sites with their appropriate scaffold. At extra-scaffold sites where NL expression was primarily diffuse, the mobile fraction of NL1 was significantly increased when compared to clusters within the same neuron (Figure 3.8 A,C, 78.2±5.8% versus 44.1±3.6% for NL1, p<0.001). Similar results were obtained for NL2 (Figure 3.8 B,D, 61.8±5.3% versus 43.5±4.7% for NL2, p<0.01). The mobile fraction of the scaffold proteins was also significantly greater at these diffuse spots (Figure 3.8 E,F). The mobile pool of PSD-95 was 77.8±5.6% at diffuse sites and 39.2±3.6 at clustered sites (n=8, p<0.001), and the mobile pool of gephyrin is 84.7±5.9% at diffuse sites and 32.2±5.9 at clustered sites (n=8, p<0.001). Thus like neurotransmitter receptors, the mobile fraction of diffusely expressed NL1 and NL2, as well as the scaffolding proteins PSD-95 and gephyrin, are significantly larger compared to sites where these proteins are clustered.  123   Figure 3.8: Neuroligin FRAP recovery at sites not associated with diffuse scaffold proteins, versus sites with clustered scaffolding proteins. (A) Example of FRAP on a cluster (box 1) and a diffuse area (box 2) of APNL1. Left panel shows the pre- bleach image, right panel shows t=0. Montages of the recovery at box 1 (top lower panel) and box 2 (lower panel) are magnified. (B) Example of FRAP on a cluster (box 1) and a diffuse area (box 2) of APNL2. (C) FRAP curves showing the mobile pool of diffuse NL1 is 78.2±5.8% (n=8) and is significantly greater than clusters of NL2 in experiment matched controls (44.1±3.6%, p<0.001). (D) FRAP curves showing the mobile pool of diffuse NL2 is 61.8±5.3% (n=8) and is significantly greater than clusters of NL2 in experiment matched controls (43.5±4.7%, p<0.01). (E) FRAP curves showing the recovery of PSD-95 at these same sites. The mobile pool of PSD-95 is 77.8±5.6% at diffuse sites and 39.2±3.6 at clustered sites (n=8, p<0.001). (F) FRAP curves showing the recovery of gephyrin CFP at these same sites. The mobile pool of gephyrin is 84.7±5.9% at diffuse sites and 32.2±5.9 at clustered sites (n=8, p<0.001). Error bars indicate S.E.M.    124 3.3.5 Neuroligins membrane mobility is determined by intracellular domains  Identified members of NL family share high sequence identity, however, intracellular regions are less conserved (Ichtchenko et al., 1996). Thus, divergence within NL1 and NL2 C-terminal domains may assist in establishment of the different reported localizations or functions of neuroligin family members. This hypothesis is supported by previous data whereby over-expression of PSD-95 resulted in enhancement of neuroligins clustering at excitatory contacts at the expense of inhibitory synapses (Levinson et al., 2005; Prange et al., 2004).  Whether specific cytoplasmic proteins interact with NL2 to regulate its sorting to inhibitory synapses remains unknown. To further understand how C-terminal interactions affect the trafficking, and in particular the surface mobile fraction, we have generated truncated forms of NL1 and NL2 lacking C-terminal domains (Figure 3.10), and monitored their dynamics using time-lapse microscopy.    Figure 3.9: Schematic of AP-tagged NL C-terminal deletions. Signal sequence (ss), acetycholine esterase-like domain (ACD), transmembrane domain (TM), membrane proximal domain (MR), proline rich domain (PR), PDZ-binding domain (PDZ).    125 Time-lapse of surface labeled NL1 and NL2 demonstrate different cluster dynamics. Although integrated fluorescence intensities of most NL2 clusters remained at constant levels (Figure 3.11 B), a deviation of integrated fluorescence from the original value was found in NL1 clusters (yellow arrowheads, Figure 3.11 A). The relative stability of NL2 clusters could be reduced by deletion of the c-terminal (Figure 3.11 C). For quantification of these changes we calculated the co-efficient of variation (CV2) over time (Kuriu et al., 2006). Plots of relative fluorescence intensity of individual clusters against time were collected from 10 clusters per neuron, derived from 2-3 culture preparations at 7 DIV. Fluorescence intensities at time 0 hour were normalized to 1, and a relative change of cluster intensities (standard deviation) was divided by the mean fluorescence in order to calculate CV (Figure 3.11 D). This plot confirmed the observation that deletion of the entire C-terminal domain increases cluster variability, since an increase in the average CV2 (0.0051±0.0001) compared to full-length NL2 (0.0033±0.0007, p<0.05, Figure 3.11 E) was calculated. Furthermore, deletion of the membrane proximal region (MR) of NL2 caused an increase in the CV2 (0.0051±0.0010) compared to full-length NL2 (p<0.05, Figure 3.11 E). Deletion of the proline-rich domain (PR) or the PDZ-binding domain had little effect on the CV2 (0.0033±0.0006, p=1.0; 0.0026±0.0004,p=0.09 respectively). Together these data demonstrate the relative stability of NL1 clusters relative to NL2 clusters, and suggest specific cytoplasmic interactions with NL2, for instance with the MR domain but not the PDZ-binding domain, can influence the relative stability of NL2 clusters.   126   Figure 3.10: Time-lapse of surface labeled NL1 and NL2 demonstrate different cluster dynamics. Images were acquired every 10 seconds. (A) In neurons expressing APNL1 (left panel), clusters marked at time 0 (yellow arrow head), show fluorescence changes over time as demonstrated by successive line scans over the same region of dendrite (right panel). (B) In neurons expressing APNL2 (left panel), smaller fluorescence changes are demonstrated (right panel). (C) In neurons expressing APNL2∆CT (left panel), larger fluorescence changes are demonstrated (right panel) compared to full length NL2.  (D) Quantification of the coefficient of variation (CV2) over time of AP-tagged NL1, NL2, NL2∆CT, NL2∆MR, NL2∆PR, NL2∆PDZ. (E) Quantification of the average coefficient of variation (CV2). Scale bars ѽҏ5µm. Error bars indicate S.E.M .   3.3.6 The mobile pool of neuroligin-1 is dependent on PDZ interactions.  Previous studies have demonstrated the importance of PDZ-interactions in the stabilization of NL1 to glutamatergic synapses (Prange et al., 2004). In order to determine whether PDZ-interactions influence the surface mobility of NL1, and like-wise NL2, we generated AP-tagged neuroligin PDZ-domain deletion mutants (NL1PDZ and NL2PDZ), and expressed them for 2 days. Deletion of the PDZ-motif of NL1 resulted in a significant increase in the mobile fraction at Shank sites (Figure 3.12 A,C; 87.7±3.4%,) when compared to full-length neuroligin (p<0.001). This further supported by the distribution of NL1∆PDZ which appears more diffuse than full-length NL2. Deletion of the PDZ-motif of NL2 did not significantly alter its recovery at gephyrin sites  127 (Figure 3.12 B,D; 46.2±4.3%, p=0.7). These data suggest that PDZ-interactions are important for influencing the mobility of NL1 but not NL2 at the appropriate sites.    Figure 3.11: PDZ interactions influence the NL1 mobile fraction. FRAP of AP-tagged NL1∆PDZ at Shank, or NL2∆PDZ at gephyrin. (A) Representative example of a site where NL1∆PDZ and YFP-Shank1b were co-clustered at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Representative example of a site where NL2∆PDZ and CFP-gephyrin were co-clustered at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of NL1 and NL1∆PDZ at Shank sites. The mobile fraction of NL1∆PDZ is 87.7±3.4% (n=14), and is significantly greater than the mobile fraction in age matched controls (70.5±3.6%, n=30, p<0.001). (D) FRAP curves comparing the recovery of NL2 and NL2∆PDZ at Shank sites. The mobile fraction of NL2∆PDZ is 46.2±4.3% (n=24), and is not significantly different than the mobile fraction in age matched controls (47.5±3.2%, n=30, p=0.7). Scale barsѽҏңµm. Error bars indicate S.E.M .   128 3.3.7 Specific scaffold proteins influence neuroligin surface mobility  To test whether interactions with scaffolding proteins may underlie the observed changes in surface mobility seen in the C-terminal deletion mutants, we expressed scaffolding proteins found at glutamatergic synapses (namely Shank or PSD-95), or at GABAergic synapses (namely gephyrin), or at both (namely S-SCAM) for 2 days, and performed FRAP experiments on AP-tagged NL1 and NL2 at these sites.  Previous studies have shown that over-expression of PSD-95 can restrict NL1 to glutamatergic synapses (Prange et al., 2004), and can shift NL2 away from GABAergic to glutamatergic synapses (Levinson et al., 2005). In order to test whether PSD-95 is able to alter the targeting of NL1 and NL2 by changing their surface mobility, we performed FRAP on AP-tagged NL¶s at PSD-95 sites after expression for 2 days at DIV 7-8. Expression of PSD-95 significantly reduced the mobile fraction of NL1 (Figure 3.13 A,C; 46.8±5.8%, p<0.001) when compared to expression with Shank. This reduction was abolished when PSD-95 was co-expressed with NL1PDZ (Figure 3.13 C;). Thus it appears that PSD-95 is able to retain NL1 at glutamatergic synapses by decreasing the mobile pool of NL1 at these sites through a PDZ-dependent mechanism. Under similar conditions, expression of PSD-95 did not significantly reduce the mobile fraction of NL2 (Figure 3.13 B,D 43.2±3.9%), compared to age matched controls at gephyrin (47.5±3.2%, p=0.7). This lack of effect on the mobile pool of NL2 suggests that PSD-95 does not cause miss-targeting of NL2 to glutamatergic synapses by decreasing the mobility of NL2 at these sites. However, these results may be confounded by the younger age of the neurons and the smaller degree of over-expression compared to previous experiments (Levinson et al., 2005).    129   Figure 3.12: PSD-95 influences NL1 mobile pool in a PDZ-dependent manner. FRAP of AP-tagged NL1 or NL2 with PSD-95 or Shank after two-days of expression. (A) Representative example of a site where NL1 and GFP-PSD-95 were co-clustered at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Representative example of a site where NL2 and GFP-PS-95 were co-clustered at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of NL1 at Shank sites and NL1 and NL1PDZ at PSD-95 sites. The mobile fraction of NL1 is 46.8±5.8% (n=12), and is significantly less than the mobile fraction at Shank sites in age matched controls (70.5±3.6%, n=30, p<0.001). The recovery of NL1PDZ at PSD-95 is 87.6±6.6% (n=18).   (D) FRAP curves comparing the recovery of NL2 at gephyrin and PSD-95 sites. The mobile fraction of NL2ҏat PSD-95 sites is 43.2±3.9% (n=14), and is not significantly different than the mobile fraction in age matched controls at gephyrin (47.5±3.2%, n=30, p=0.7). Scale bars ѽҏңµm. Error bars indicate S.E.M .   S-SCAM expression occurs at both glutamatergic and GABAergic synapses, and can interact with NL1 in a PDZ-dependent manner (Iida et al., 2004) and NL2 in a PDZ- independent manner (Sumita et al., 2007). Expression of NLs with S-SCAM lead to a dramatic decrease in the mobile fraction of NL1 (Figure 3.14 A,C; 22.9±2.9%, p<0.001), and NL2 (Figure 3.14 B,D; 33.3±4.1%, p<0.001). For both NL1 and NL2, the mobile fraction was well-correlated with the mobile fraction of S-SCAM (r=0.46, p<0.05 for NL1, and r=0.92, p<0.001 for NL2).    130  Figure 3.13: S-SCAM influences the mobile fraction of NL1 and NL2. (A) Representative example of a site where AP-NL1 and GFP S-SCAM were co-clustered at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) Site where AP- NL2 and GFP S-SCAM  was co-clustered (left panels) were bleached and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves comparing the recovery of NL1 at Shank versus S- SCAM. The mobile fraction of APNL1 was 22.9±2.4% (n=22), and significantly decreased compared to NL1 at Shank sites in age-matched controls (p<0.001). (D) FRAP curves comparing the recovery of NL2 at gephyrin versus S-SCAM. The mobile fraction of NL2 was 33.3±4.1% (n=20), and significantly decreased compared to NL2 at gephyrin sites in age-matched controls (p<0.001). Scale barsѽҏңµm. Error bars indicate S.E.M .   These results showing a change in the size of the mobile fraction are in agreement with the enhanced clustering of NL1 and NL2 upon co-expression of S-SCAM (Figure 3.15 A,B).  Exogenous expression of NL1 or NL2 for 3 days leads to a primarily diffuse expression along the entire neuron membrane with few distinct clusters (Figure 3.15 C, black bars: 0.58±0.04 and 0.56±0.04 clusters per µm respectively). Co-expression with S- SCAM leads to increased clustering of NL1 (0.81±0.05, p<0.01), and NL2 (0.73±0.04, p<0.01). Furthermore, the ability of S-SCAM to cluster and decrease the mobile fraction NLs appears to influence the function of NL in its ability to enhance presynaptic terminals. Enhancement of the average VGlut intensity (compared to untransfected controls within the same field of view) in presynaptic terminals apposing onto neurons expressing NL and S-SCAM was increased when compared to neurons transfected with NL alone (Figure 3.15 D). NL1 alone increased VGlut1 intensity 1.82±0.10 fold, and was significantly potentiated by co-expression of GFP S-SCAM to 2.18±0.13 fold (p<0.05).  131 NL2 alone increased VGlut1 intensity 2.34±0.19 fold, and was significantly potentiated by co-expression of GFP S-SCAM to 3.15±0.32 fold (p<0.05). Together these data suggest that S-SCAM is a mediator of NL retention at both glutamatergic and GABAergic sites. However, due to its location at both of these types of synapses and its ability to influence the clustering of both NL1 and NL2, the role of S-SCAM may be as a general scaffold µtrap¶, rather than a mechanism of sorting to specific synapse sites. This hypothesis is supported by the fact that co-expression of NL2 and S-SCAM can further enhance VGlut terminals.    Figure 3.14: S-SCAM enhances clustering and function of neuroligins. Neurons were transfected with HA-tagged NL1 or NL2 alone or co-transfected with GFP S-SCAM from DIV7-10 in order to assess the ability of S-SCAM to cluster NL, and modulate the presynaptic enhancement of VGlut caused by NL. (A) Expression of NL1 is primarily diffuse (left panels), whereas co- expression with GFP S-SCAM increases NL1 clustering (right panels). (B) Expression of NL2 is primarily diffuse (left panels), whereas co-expression with GFP S-SCAM increases NL2 clustering (right panels). (C) Quantification of the clustering of NL1 and NL2 by S-SCAM. The number of NL1 clusters per µm increased from 0.58±0.04 to 0.81±0.05 when co-expressed with GFP S-SCAM (p<0.01). The number of NL2 clusters increased from 0.56±0.04 to 0.73±0.04 when co-expressed with S-SCAM (p<0.05).  (D) Enhancement of presynaptic maturation as assessed by intensity of apposed Vglut1 clusters on transfected neurons, normalized to VGlut1 intensity of un-transfected controls within the same field of view. Quantification shows that GFP S-SCAM can increase VGlut1 intensity 1.36±0.08 fold compared to un- transfected controls. NL1 alone increased VGlut1 intensity 1.82±0.10 fold, and was significantly potentiated by co-expression of GFP S-SCAM to 2.18±0.13 fold (p<0.05). NL2 alone increased VGlut1 intensity 2.34±0.19 fold, and was significantly potentiated by co-expression of GFP S-SCAM to 3.15±0.32 fold (p<0.05). Scale barsѽҏ5 µm. Error bars indicate S.E.M .    132 3.3.8 Neuroligin mobility is differentially influenced by the actin and microtubule cytoskeleton.  Previous studies have indicated that basic elemental proteins required to assemble and retain at young hippocampal synapses made during the first week in culture, are reliant on their links to the actin cytoskeleton (Zhang and Benson, 2001), and that over time and as synapses incorporate additional proteins, maintenance of synapse ultra-structure and the localization of many synaptic proteins become independent of F-actin and microtubules (Allison et al., 2000). To test whether neuroligin mobility at the surface is influenced by the actin or microtubule cytoskeleton we performed FRAP in the presence of 3µM nocodozole (NOCOD) or 4µM cytochalasin B (CytoB), pharmacological agents known to disrupt the microtubule and actin cytoskeleton respectively, through filament de- polymerization (Sabo and McAllister, 2003). This treatment did not affect the overall morphology of the neurons within the time period of the FRAP experiments (Appendix Figure 5.10). De-polymerization of microtubules moderately decreased the mobile pool of NL1 (54.2±9.5%, p<0.05), whereas blocking the re-polymerization of the actin cytoskeleton dramatically decreased NL1 recovery (28.2±4.2%, p<0.001; Figure 3.16 A,C).  Blocking the re-polymerization of either actin or microtubules had the similar effect on NL2: both significantly reduced the mobile pool (Figure 3.16 B,C; 25.8±2.9%, p<0.001 with CytoB; 25.1±2.8%, p<0.01 with NOCOD).  Filamentous actin has a critical role in determining the extent of dynamic re-organization in PSD molecular composition (Kuriu et al., 2006). In agreement with these studies, we find that the mobile fraction of Shank is dramatically reduced in the presence of CytoB (Figure 3.17; 34±4.1%, p<0.001), and in addition, that the mobile fraction of Shank is also reduced in the presence of NOCOD (54.0±3.9%, p<0.05), albeit to a lesser degree. In contrast, the size of the mobile pool of gephyrin was only influenced by CytoB (Figure 3.17; 32.8±5.2%). Thus, not only are scaffolds at the glutamatergic synapses influenced by both actin, but they are also influenced by microtubule re-arrangement. In addition, the role of actin re-arrangement had been extended to GABAergic scaffold proteins.  133   Figure 3.15: Neuroligin surface mobility is differentially influenced by the actin and microtubule cytoskeleton. (A) AP-NL1 and GFP Shank co-cluster in presence of CytoB at DIV7 (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) AP-NL2 and CFP gephyrin co-cluster in presence of CytoB (left panels), bleached and the fluorescence recovery monitored by time-lapse (right panel). (C) APNL1 and YFP-Shank1b co-cluster in the presence of NOCOD (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (D) APNL2 and CFP-gephyrin co-cluster in the presence of NOCOD (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (E) FRAP curves comparing APNL1 at Shank in the presence of NOCOD and CytoB. The mobile fraction of NL1 was 54.2±9.5% (n=14) in the presence of NOCOD, and 28.2±4.2% (n=20) in the presence of CytoB. Both treatments significantly decreased the mobile fraction of NL1 compared to untreated NL1 at Shank sites in age-matched controls (p<0.05, and p<0.001 respectively). (F) FRAP curves of APNL2 at gephyrin in the presence of NOCOD and CytoB. The mobile fraction of NL2 was 25.1±2.8% (n=20) in the presence of NOCOD, and 25.8±2.9% (n=18) in the presence of CytoB. Both treatments significantly decreased the mobile fraction of NL1 compared to untreated NL1 at Shank sites in age- matched controls (p<0.001 and p<0.01 respectively). Scale bars ѽҏңµm. Error bars indicate S.E.M .     134  Figure 3.16: Scaffolding proteins are differentially regulated by the actin and microtubule cytoskeleton. (A) YFP-Shank1b at DIV7 in the presence of CytoB (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (B) YFP-Shank1b at DIV7 in the presence of NOCOD (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (C) FRAP curves of GFP Shank in the presence of NOCOD and CytoB. The mobile fraction of Shank was 54.0±3.9%  (n=14) in the presence of NOCOD, and 34±4.1% (n=20) in the presence of CytoB. Both treatments significantly decreased the mobile fraction of Shank compared to untreated Shank clusters in age-matched controls (p<0.05 and p<0.001 respectively). (D) CFP-gephyrin at DIV7 in the presence of CytoB (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (E) CFP-gephyrin at DIV7 in the presence of NOCOD (left panels), bleached, and the fluorescence recovery monitored by time-lapse (right panel). (F) FRAP curves of CFP gephyrin in the presence of NOCOD and CytoB. The mobile fraction of gephyrin was 53.5±6.5%  (n=20) in the presence of NOCOD, and 32.8±5.2%  (n=18) in the presence of CytoB. Only CytoB treatment significantly decreased the mobile fraction of gephyrin compared to untreated gephyrin sites in age-matched controls (p<0.001). Scale barsѽҏңµm. Error bars indicate S.E.M., and n represents the number of clusters analyzed from 2 independent experiments.     135 3.3.9 Neuroligin-2 surface retention is determined by intracellular domains  In addition to the horizontal trafficking, receptors at glutamatergic and GABAergic under go vertical trafficking into and out of the membrane surface. In order to test the vertical forward trafficking of neuroligins into the membrane we performed pulse-chase experiments on neurons transfected with AP-tagged NL2 or NL2 c-terminal deletion mutants, by sequential incubation with mSA conjugated to Alexa568 and followed by Alexa488 24 hour later. The average intensity of newly inserted NL was measured over 3-5 dendrite sections of at least 100 microns for each neuron. These values were averaged and then compared to NL2 full-length. All four C-terminal deletion mutants showed comparable insertion under these conditions, albeit slightly reduced compared to full- length NL2 (Figure 3.18 A,B). The average intensity of inserted NL2 was 154.9±16.4 (AU) and was not significantly different than NL2 ∆CT (138.3±29.2, p=0.6), NL2 ∆MR (140.0±16.3, p=0.5), NL2 ∆PDZ (129.4±20.2, p=0.3), or NL2 ∆PR (135.3±33.1, p=0.6). These data suggest that c-terminal interactions are not important for the insertion of NL2.   Figure 3.17: Insertion of AP-tagged neuroligin-2. (A) Representative images taken under similar imaging conditions of newly inserted AP-tagged NL2 and NL2 mutants. (B) Quantification of the average intensity of newly inserted NL2. Scale bars ѽҏңµm. Error bars indicate S.E.M., and n represents the number of neurons analyzed (8-15), from 2 independent experiments.    In order to test the inward horizontal trafficking of NL2, we incubated AP-tagged NL2 or NL2 c-terminal deletion mutants transfected neurons with mSA Alexa568, and 24 hours later assayed the amount of mSA internalized by its resistance to trypan blue quenching (Harata et al., 2006). Neurons were co-transfected with YFP-tagged BirA KDEL which  136 filled the endoplasmic reticulum and allowed for visualization of dendrites. Images of the mSA Alexa568 were taken under identical conditions before (Figure 3.19 A,B; left panels) and after treatment with 0.5% trypan blue (right panels). The ratio of the average intensity of unquenched signal after treatment, compared to before treatment was measured over the cell body and dendrites (Figure 3.19 C). This analysis revealed that NL2∆CT and NL2∆MR had a significantly larger unquenched signal within the cell body (0.79±0.03 and 0.84±0.03 respectively, p<0.001), and dendrites (0.64±0.06 and 0.64±0.04 respectively, p<0.001) when compared to full-length NL2 (0.58±0.04 for the cell body, 0.40±0.05 for the dendrites). This data suggests that the amount of internalized NL was greater for these mutants. In contrast, the amount of unquenchable signal on the cell body and dendrites in neurons transfected with NL2∆PDZ (0.55±0.07, p=0.7 and 0.39±0.08, p=0.9) or NL2∆PR (0.50±0.07, p=0.4 and 0.42±0.09, p=0.8) was comparable to full-length NL2. These data suggest that specific domains are important in maintaining NL2 on the surface.   Figure 3.18: Internalization of neuroligin-2 is dependent on C-terminal domains. (A) Representative image of AP-tagged NL2, 24 hours after labeling with mSA Alexa 568 (left panel). The same neuron after treatment with quenching agent trypan blue (right panel). (B) Representative image of AP-tagged NL2∆CT, 24 hours after labeling with mSA Alexa 568 (left panel). The same neuron after treatment with quenching agent trypan blue (right panel).  (C) Quantification of the unquenched signal where initial average intensity is normalized to 1. Scale barsѽҏ5µm. Error bars indicate S.E.M.   137 3.4 DISCUSSION  Signals received through excitatory versus inhibitory contacts control neuronal excitability and brain function, however, the mechanisms that regulate the formation of the diverse types of synapses in the CNS are largely unknown. The goal of this work was to clarify mechanisms that may regulate the formation of a synapses by defining factors that sort proteins to specific synapse types (excitatory versus inhibitory) in vitro, in order to better understand the function of these proteins in vivo. The importance of the development of proper synaptic balance is highlighted when looking at instances where this is perturbed, as seen in mental diseases such as autism (Bartlett et al., 2005; Belmonte et al., 2004; Jamain et al., 2003; Laumonnier et al., 2004; Rubenstein and Merzenich, 2003; Talebizadeh et al., 2004; Yan et al., 2004; Ylisaukko-oja et al., 2005). Here we demonstrate that NLs can freely diffuse between postsynaptic and extra- postsynaptic areas (Figure 3.1)., and this mobile fraction was determined to be primarily composed of proteins found on the surface (Figure 3.3). Factors that could modulate this trafficking include developmental age (Figure 3.2 and 3.4), intracellular interactions with specific scaffolding proteins (Figure 3.12 and 3.13), and the cytoskeleton (Figure 3.15). These results provide novel insight into the trafficking and processes that may underlie NL retention and function at the appropriate site.  Being a transmembrane protein, neuroligin trafficking could occur through direct insertion into postsynaptic membranes, or through extra-synaptic insertion followed by lateral diffusion. Such a hypothesis is similar to what has been proposed for neurotransmitter receptors at both glutamatergic and GABAergic synapses (Choquet and Triller, 2003; Triller and Choquet, 2005). Previous results looking at the adhesion moelcule L1 in neuronal growth cones (Dequidt et al., 2007), and the adhesion molecule NCAM  during cell migration and neurite outgrowth (Conchonaud et al., 2007), have suggested the importance of lateral diffusion in the trafficking of adhesion molecules. Whether lateral diffusion is important in the trafficking of other adhesion complexes is unknown, and it is still unclear how these complexes are trafficked to synaptic sites. Our results support the trafficking and retention of neuroligins to synaptic sites through lateral  138 diffusion. Not only do clusters of NL1 or NL2 have a mobile fraction which can be measured by flourescence recovery after photobleaching (FRAP), but when we use a technique that specifically labels the surface population of NL1 or NL2, the FRAP dynamics are similar. Thus it appears that neuroligin trafficking to the appropriate sites occurs through the trapping of diffusing NL. This result may explain why endoengous staining of neuroligins shows weak clusters with diffuse staining nonsynaptically, and then appear more distinctly clustered at later developmental stages (Gerrow, 2005; Levinson et al., 2005). Indeed, our results show that the mobile fraction of NL1 and NL2 are reduced at later ages, suggesting that these protein are more stably associated with scaffold sites, and agrees with previous studies that demonstrate that general characteristic of membrane proteins is that their movement into and out of spines is slow compared with that in non-spiny membrane (Ashby et al., 2006). How is the specificity of NL isoforms for glutamatergic and GABAergic synapses encoded? Two possibilities arise: specificity could arise from selective extracellular interactions between specific NL and NRX isoforms; or NL isoforms could be recruited to glutamatergic versus GABAergic sites by interaction with synapse-specific cytoplasmic scaffolds. Data thus far suggests an inter-play between both possibilities. In support of extracellular interactions determining specificity, previous studies have shown that alternative splicing of the NL extracellular domain can control their interaction with specific β-NRX splice variants, as well as their differential localization and synapse- inducing activity toward glutamatergic versus GABAergic axons (Chih et al., 2005). Furthermore, in COS cell neuron co-culture assays, α-NRX induce clustering of GABAergic postsynaptic scaffolding proteins, such as gephyrin and NL 2, but not of the glutamatergic postsynaptic scaffolding protein PSD-95 or NL 1/3/4 (Kang et al., 2007). In support for a cytoplasmic specificity mechanism, over-expression of PSD-95, a glutamatergic scaffolding protein, restricts NL1 to glutamatergic sites, and can recruit endogenous NL2 to glutamatergic synapses (Graf et al., 2004; Levinson and El-Husseini, 2005; Prange et al., 2004). Here, we further demonstrate that importance of intracellular interactions, and show that they act by influencing the mobile fraction of NLs. For NL1, PDZ-dependent interactions are important, as supported by the increase in the mobile  139 fraction in PDZ-deletion mutants, but also the decrease in the mobile fraction with the increased presence of PSD-95 or S-SCAM. Interestingly, there was little effect by another PDZ protein Shank, suggesting specificity for certain proteins to influence the surface mobility of NLs. This does not infer that Shank does not influence NL trafficking per se, since Shank may exert effects that do not influence the mobile fraction at the developmental stages than studied herein. Since Shank has a prominent role in spine morphogenesis, it may be hypothesized that Shank may exert a larger effect on neuroligin during these later stages of synapse maturation. For NL2, PDZ-interactions were not important in influencing the mobile fraction, as supported by the lack of influence on the deletion of the PDZ-domain. As with NL1, the expression of specific scaffolding proteins was able to influence the mobile fraction of NL2, namely S-SCAM. These results for intracellular mechanisms for limiting the mobile fraction of NL1 and NL2 are summarized in the schematic diagram in Figure 3.22.  Figure 3.19: Schematic of interactions that influence the lateral mobility of neuroligins. Diagram of the interactions predicted to affect the lateral trafficking of NL1 at glutamatergic (left panel), and the interactions predicted to affect the lateral trafficking of NL2 at GABAergic synapses (right panel).    140 What are the other potential intracellular interacting proteins for neuroligin-2 at inhibitory synapses? The GRIP family of proteins (GRIP1 and GRIP2 also known as ABP; AMPA receptor binding protein) represent another important family of scaffolding proteins, which associate with subunits of the AMPA receptor at glutamatergic synapses (O'Brien et al., 1998). However, other studies showed that GRIP is also present at inhibitory synapses (Charych et al., 2004).  Thus, members of the GRIP family represent a class of scaffolding proteins potentially involved in the development of both excitatory and inhibitory contacts. Although it is unknown whether GRIP directly interacts with NL2 at inhibitory synapses, if GRIP does influence NL2 stability at inhibitory synapses, our results indicate that this may not be through a PDZ-dependent interaction, since deletion of the PDZ-binding domain of NL2 had little effect on its stability as measured by time-lapse and FRAP. Then what is the purpose of the PDZ-binding domain of NL2 if it is not required for its stability at the appropriate site? Perhaps interaction with PDZ proteins modulates the recruitment of NL2 away from inhibitory synapses as a mechanism to modulate maturation. NL1 is primarily at glutamatergic and NL2 is primarily at GABAergic synapses in hippocampal neurons, and a fraction of each can be found at the 'wrong' synapse (Chih et al., 2005; Levinson et al., 2005). What about other NL isoforms? Interestingly, NL3 can be found at both glutamatergic and GABAergic synapses (Budreck and Scheiffele, 2007). Individual synapses contained both NL2 and NL3, and co-immunoprecipitation studies revealed the presence of NL1/3 and NL2/3 complexes in brain extracts, suggesting that neruroligin-3 is a shared component of glutamatergic and GABAergic synapses. Thus, in addition to extracellular and cytoplasmic interactions that can influence NL targeting, factors that influence NL dimerization may also contribute to neuroligin targeting.  In addition to scaffolding molecules, the actin and microtubule cytoskeleton can structurally organize synapses.  Here, we described a novel link between cytoskeleton depolymerization and the amount of mobile NLs on the surface. We found that the mobility of NL1 was decreased by microtubule depolymerization and dramatically decreased by actin depolymerization. Interestingly, the observed decrease in the mobile  141 fraction paralleled a decrease in the mobile fraction of Shank at these sites. The observation that amount of mobile Shank is decreased with actin de-polymerization has previously been observed for other scaffolding proteins, namely PSD-95, GKAP, and PSD-ZIP45 (Kuriu et al., 2006). Although the trafficking of these proteins has been shown to less effected by cytoskeleton depolymerization in older neuronal cultures (Allison et al., 2000), and may be due to the presence of additional protein interactions that are predominant in syapses of older neurons, but absent in younger neurons.  Thus, it appears that actin depolymerization has a global effect on the mobile fraction of many of proteins at young glutamatergic synapses. These results are not surprising given the multitude of protein interactions with the actin cytoskeleton and the PSD. However, the actin and microtubule cytoskeleton may also influence the mobility of these proteins between PSD sites. Indeed, at the surface of neuronal growth cones or fibroblast lamellipodia, both of which bear a particularly active cytoskeleton, rearward actin movement is able to transport surface glycoproteins (Sheetz et al., 1989). Such transport has been shown for several adhesion proteins involved in cell migration such as integrins (Felsenfeld et al., 1996; Grabham et al., 2000), neuron-glia related cell-adhesion molecule (NrCAM) (Faivre-Sarrailh et al., 1999), aplysia cell-adhesion molecule (ApCAM) (Thompson et al., 1996) and cadherins (Lambert et al., 2002). This mobility between PSD sites may explain the interaction between NL1 and microtubule cytoskeleton, since microtubules are largely absent from the PSD. A similar mechanism may be influencing the diffusion of NL1 between scaffold sites.  The actin and microtubule cytoskeleton has also been shown to influence proteins at GABAergic synapses. Previous studies have shown that treatment with cytochalasin D resulted in a reduction in the number of gephyrin clusters (Bausen et al., 2006). This reduction in gephyrin clusters may be precluded by the observed reduction in the mobile pool that we observe, where decreased mobility may inhibit the formation of new clusters. Several molecular links between gephyrin and the microtubular cytoskeleton have already been identified (Giesemann et al., 2003)(Kirsch et al., 1991; Ramming et al.,  142 2000) (Fuhrmann et al., 2002). Therefore, our finding that the mobile fraction of gephyrin is relatively unchanged when microtubules are depolymerized is surprising. Our results are different from the disruption of clusters of gephyrin in spinal neurons by the microtubule depolymerizing agent demecolcine (Kirsch and Betz, 1995), however these differences may be due to the synapse and cell type used.  Although studies have linked scaffolding molecules and receptors of GABAergic synapses to the turn-over the cytoskeleton, it is unclear how adhesion complexes are influenced. Our study reveals an important role for both the actin and microtubule cytoskeleton in the lateral mobility of NL2 at gephyrin sites. Could this interaction help determine its proper localization? Interestingly, f-actin-dependent synaptic localization of nectin-1, a immunoglobulin-like adhesion molecules, and it binding partner l-afadin which interacts with a number of other proteins such as the Ras family GTPases, are present at nascent excitatory and inhibitory synapses, but with maturation nectin-1 is lost at inhibitory synapses. The early synaptic localization of nectin-1 and l-afadin is dependent on F-actin, however, at later stages the localization of nectin-1 becomes independent of F-actin which co-relates with its loss at inhibitory synapses (Lim et al., 2008). Future studies will be needed to determine if a similar mechanism of actin-dependence could also assist in the proper targeting of NL2.  Recent studies have found that the endocytosis of postsynaptic cargo occurs at endocytic zones (EZs), stably positioned sites of clathrin assembly adjacent to the PSD (Blanpied et al., 2002; Racz et al., 2004). The tight localization of postsynaptic endocytosis is thought to control postsynaptic membrane composition and regulate diverse aspects of synaptic transmission and plasticity (Kennedy and Ehlers, 2006). We show that c-terminal domains are important for the surface stability of NL2, which may in turn influence its function. Similar results have been demonstrated for NMDA receptors, where a clathrin- dependent internalization motif found on the distal C-terminus of NR2B which is negatively regulated by intracellular interactions with PSD-95 (Lavezzari et al., 2003; Roche et al., 2001). The interaction(s) that modulate NL2 stability on the surface are unknown. However, these interactions may be modulated by activity since the surface  143 localization another adhesion molecule, OBCAM , is dynamically regulated in response to neuronal activity (Yamada et al., 2007). Future studies with activity paradigms may uncover a similar mechanism controlling the function of NLs.  Together these results describe the surface trafficking of NLs, and determine several of the intracellular mechanisms that can influence the mobility of the NL1 at Shank sites, and NL2 at gephyrin sites. These results provide insight into the processes that may underlie their retention at the appropriate site, and thus, modulate NL function.  144 3.5 REFERENCES  Allison, D. W., Chervin, A. S., Gelfand, V. I., and Craig, A. M. (2000). Postsynaptic Scaffolds of Excitatory and Inhibitory Synapses in Hippocampal Neurons: Maintenance of Core Components Independent of Actin Filaments and Microtubules. J Neurosci 20, 4545-4554.  Ashby, M. C., Maier, S. R., Nishimune, A., and Henley, J. M. (2006). Lateral Diffusion Drives Constitutive Exchange of AMPA Receptors at Dendritic Spines and Is Regulated by Spine Morphology. J Neurosci 26, 7046-7055.  Axelrod, D., Koppel, D. E., Schlessinger, J., Elson, E., and Webb, W. W. (1976). Mobility measurement by analysis of fluorescence photobleaching recovery kinetics. Biophys J 16, 1055-1069.  Bartlett, C. W., Gharani, N., Millonig, J. H., and Brzustowicz, L. M. (2005). Three autism candidate genes: a synthesis of human genetic analysis with other disciplines. Int J Dev Neurosci 23, 221-234.  Bausen, M., Fuhrmann, J. C., Betz, H., and O'Sullivan, G. A. (2006). The state of the actin cytoskeleton determines its association with gephyrin: Role of ena/VASP family members. Molecular and Cellular Neuroscience 31, 376-386.  Belmonte, M. K., Allen, G., Beckel-Mitchener, A., Boulanger, L. M., Carper, R. A., and Webb, S. J. (2004). Autism and abnormal development of brain connectivity. J Neurosci 24, 9228-9231.  Blanpied, T. A., Scott, D. B., and Ehlers, M. D. (2002). Dynamics and regulation of clathrin coats at specialized endocytic zones of dendrites and spines. Neuron 36, 435-449.  Bresler, T., Ramati, Y., Zamorano, P. L., Zhai, R., Garner, C. C., and Ziv, N. E. (2001). The dynamics of SAP90/PSD-95 recruitment to new synaptic junctions. Mol Cell Neurosci 18, 149-167.  Bresler, T., Shapira, M., Boeckers, T., Dresbach, T., Futter, M., Garner, C. C., Rosenblum, K., Gundelfinger, E. D., and Ziv, N. E. (2004). Postsynaptic density assembly is fundamentally different from presynaptic active zone assembly. J Neurosci 24, 1507-1520.  Budreck, E. C., and Scheiffele, P. (2007). Neuroligin-3 is a neuronal adhesion protein at GABAergic and glutamatergic synapses. European Journal of Neuroscience 26, 1738-1748.  Charych, E. I., Yu, W., Li, R., Serwanski, D. R., Miralles, C. P., Li, X., Yang, B. Y., Pinal, N., Walikonis, R., and De Blas, A. L. (2004). A four PDZ domain-containing splice variant form of GRIP1 is localized in GABAergic and glutamatergic synapses in the brain. J Biol Chem 279, 38978-38990.  Chih, B., Engelman, H., and Scheiffele, P. (2005). Control of Excitatory and Inhibitory Synapse Formation by Neuroligins. Science 307, 1324-1328.  Choquet, D., and Triller, A. (2003). The role of receptor diffusion in the organization of the postsynaptic membrane. 4, 251-265.  Conchonaud, F., Nicolas, S., Amoureux, M.-C., Menager, C., Marguet, D., Lenne, P.-F., Rougon, G., and Matarazzo, V. (2007). Polysialylation Increases Lateral Diffusion of Neural Cell Adhesion Molecule in the Cell Membrane. J Biol Chem 282, 26266-26274.  Deng, L., Yao, J., Fang, C., Dong, N., Luscher, B., and Chen, G. (2007). Sequential Postsynaptic Maturation Governs the Temporal Order of GABAergic and Glutamatergic Synaptogenesis in Rat Embryonic Cultures. J Neurosci 27, 10860-10869.  145  Dequidt, C., Danglot, L., Alberts, P., Galli, T., Choquet, D., and Thoumine, O. (2007). Fast Turnover of L1 Adhesions in Neuronal Growth Cones Involving Both Surface Diffusion and Exo/Endocytosis of L1 Molecules. Mol Biol Cell 18, 3131-3143.  Essrich, C., Lorez, M., Benson, J. A., Fritschy, J.-M., and Luscher, B. (1998). Postsynaptic clustering of major GABAA receptor subtypes requires the [gamma]2 subunit and gephyrin. 1, 563-571.  Faivre-Sarrailh, C., Falk, J., Pollerberg, E., Schachner, M., and Rougon, G. (1999). NrCAM, cerebellar granule cell receptor for the neuronal adhesion molecule F3, displays an actin-dependent mobility in growth cones. J Cell Sci 112 Pt 18, 3015-3027.  Felsenfeld, D. P., Choquet, D., and Sheetz, M. P. (1996). Ligand binding regulates the directed movement of beta1 integrins on fibroblasts. Nature 383, 438-440.  Feng, G., Tintrup, H., Kirsch, J., Nichol, M. C., Kuhse, J., Betz, H., and Sanes, J. R. (1998). Dual requirement for gephyrin in glycine receptor clustering and molybdoenzyme activity. Science 282, 1321- 1324.  Fu, Z., Washbourne, P., Ortinski, P., and Vicini, S. (2003). Functional excitatory synapses in HEK293 cells expressing neuroligin and glutamate receptors. J Neurophysiol 90, 3950-3957.  Gerrow, K., Romorini, S., Nabi, S.,  Colicos, M., Sala, C.; El-Husseini, A. (2005). A preformed complex of postsynaptic proteins is involved in excitatory synapse development. Neuron 49, 547-562.  Giesemann, T., Schwarz, G., Nawrotzki, R., Berhorster, K., Rothkegel, M., Schluter, K., Schrader, N., Schindelin, H., Mendel, R. R., Kirsch, J., and Jockusch, B. M. (2003). Complex formation between the postsynaptic scaffolding protein gephyrin, profilin, and Mena: a possible link to the microfilament system. J Neurosci 23, 8330-8339.  Grabham, P. W., Foley, M., Umeojiako, A., and Goldberg, D. J. (2000). Nerve growth factor stimulates coupling of beta1 integrin to distinct transport mechanisms in the filopodia of growth cones. J Cell Sci 113 ( Pt 17), 3003-3012.  Graf, E. R., Zhang, X., Jin, S. X., Linhoff, M. W., and Craig, A. M. (2004). Neurexins Induce Differentiation of GABA and Glutamate Postsynaptic Specializations via Neuroligins. Cell 119, 1013- 1026.  Harata, N., Choi, S., Pyle, J.L., Aravanis, A.M., and Tsien, R.W. (2006). Frequency dependent kinetics and prevailence of kiss-and-run and reuse at hippocampal synapses studied with novel quenching methods. Neuron. 49, 243-256.  Howarth, M., Chinnapen, D. J., Gerrow, K., Dorrestein, P. C., Grandy, M. R., Kelleher, N. L., El-Husseini, A., and Ting, A. Y. (2006). A monovalent streptavidin with a single femtomolar biotin binding site. Nat Methods 3, 267-273.  Ichtchenko, K., Nguyen, T., and Südhof, T. C. (1996). Structures, Alternative Splicing, and Neurexin Binding of Multiple Neuroligins. J Biol Chem 271, 2676-2682.  Iida, J., Hirabayashi, S., Sato, Y., and Hata, Y. (2004). Synaptic scaffolding molecule is involved in the synaptic clustering of neuroligin. Mol Cell Neurosci 27, 497-508.  Jacob, T. C., Bogdanov, Y. D., Magnus, C., Saliba, R. S., Kittler, J. T., Haydon, P. G., and Moss, S. J. (2005). Gephyrin Regulates the Cell Surface Dynamics of Synaptic GABAA Receptors. J Neurosci 25, 10469-10478.   146 Jamain, S., Quach, H., Betancur, C., Rastam, M., Colineaux, C., Gillberg, I. C., Soderstrom, H., Giros, B., Leboyer, M., Gillberg, C., et al. (2003). Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism. Nat Genet.  Kang, Y., Zhang, X., Dobie, F., Wu, H., and Craig, A. M. (2007). Induction of gabaergic postsynaptic differentiation by alpha -neurexins. J Biol Chem, M703957200.  Kennedy, M. J., and Ehlers, M. D. (2006). Organelles and trafficking machinery for postsynaptic plasticity. Annu Rev Neurosci 29, 325-362.  Kirsch, J., and Betz, H. (1995). The postsynaptic localization of the glycine receptor-associated protein gephyrin is regulated by the cytoskeleton. J Neurosci 15, 4148-4156.  Kneussel, M., and Betz, H. (2000). Clustering of inhibitory neurotransmitter receptors at developing postsynaptic sites: the membrane activation model. Trends Neurosci 23, 429-435.  Kneussel, M., Brandstatter, J. H., Gasnier, B., Feng, G., Sanes, J. R., and Betz, H. (2001). Gephyrin- independent clustering of postsynaptic GABA(A) receptor subtypes. Mol Cell Neurosci 17, 973-982.  Kneussel, M., Brandstatter, J. H., Laube, B., Stahl, S., Muller, U., and Betz, H. (1999). Loss of postsynaptic GABA(A) receptor clustering in gephyrin-deficient mice. J Neurosci 19, 9289-9297.  Kuriu, T., Inoue, A., Bito, H., Sobue, K., and Okabe, S. (2006). Differential Control of Postsynaptic Density Scaffolds via Actin-Dependent and -Independent Mechanisms. J Neurosci 26, 7693-7706.  Lambert, M., Choquet, D., and Mege, R. M. (2002). Dynamics of ligand-induced, Rac1-dependent anchoring of cadherins to the actin cytoskeleton. J Cell Biol 157, 469-479.  Laumonnier, F., Bonnet-Brilhault, F., Gomot, M., Blanc, R., David, A., Moizard, M. P., Raynaud, M., Ronce, N., Lemonnier, E., Calvas, P., et al. (2004). X-Linked Mental Retardation and Autism Are Associated with a Mutation in the NLGN4 Gene, a Member of the Neuroligin Family. Am J Hum Genet 74, 552-557.  Lavezzari, G., McCallum, J., Lee, R., and Roche, K. W. (2003). Differential binding of the AP-2 adaptor complex and PSD-95 to the C-terminus of the NMDA receptor subunit NR2B regulates surface expression. Neuropharmacology Post-translational modifications of protein structure and synaptic function 45, 729- 737.  Levi, S., Logan, S. M., Tovar, K. R., and Craig, A. M. (2004). Gephyrin is critical for glycine receptor clustering but not for the formation of functional GABAergic synapses in hippocampal neurons. J Neurosci 24, 207-217.  Levinson, J. N., Chery, N., Huang, K., Wong, T. P., Gerrow, K., Kang, R., Prange, O., Wang, Y. T., and El-Husseini, A. (2005). Neuroligins Mediate Excitatory and Inhibitory Synapse Formation: INVOLVEMENT OF PSD-95 AND NEUREXIN-1{beta} IN NEUROLIGIN-INDUCED SYNAPTIC SPECIFICITY. J Biol Chem 280, 17312-17319.  Levinson, J. N., and El-Husseini, A. (2005). Building Excitatory and Inhibitory Synapses: Balancing Neuroligin Partnerships. Neuron 48, 171-174.  Lim, S. T., Lim, K. C., Giuliano, R. E., and Federoff, H. J. (2008). Temporal and spatial localization of nectin-1 and l-afadin during synaptogenesis in hippocampal neurons. J Comp Neurol 507, 1228-1244.  Lu , J., Helton, T. D., Blanpied, T. A., Rácz, B., Newpher, T. M., Weinberg, R. J., and Ehlers, M. D. (2007). Postsynaptic positioning of endocytic zones and AMPA receptor cycling by physical coupling of dynamin-3 to Homer. Neuron 55, 874-889.  147  McAllister, A. K. (2007). Dynamic Aspects of CNS Synapse Formation. Annual Review of Neuroscience 30, 425-450.  O'Brien, R. J., Lau, L. F., and Huganir, R. L. (1998). Molecular mechanisms of glutamate receptor clustering at excitatory synapses. Curr Opin Neurobiol 8, 364-369.  Okabe, S., Kim, H. D., Miwa, A., Kuriu, T., and Okado, H. (1999). Continual remodeling of postsynaptic density and its regulation by synaptic activity. Nat Neurosci 2, 804-811. Prange, O., Wong, T. P., Gerrow, K., Wang, Y. T., and El-Husseini, A. (2004). A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. PNAS 101, 13915-13920.  Racz, B., Blanpied, T. A., Ehlers, M. D., and Weinberg, R. J. (2004). Lateral organization of endocytic machinery in dendritic spines. Nat Neurosci 7, 917-918.  Roche, K. W., Standley, S., McCallum, J., Dune Ly, C., Ehlers, M. D., and Wenthold, R. J. (2001). Molecular determinants of NMDA receptor internalization. 4, 794-802.  Romorini, S., Piccoli, G., Jiang, M., Grossano, P., Tonna, N., Passafaro, M., Zhang, M., and Sala, C. (2004). A functional role of postsynaptic density-95-guanylate kinase-associated protein complex in regulating Shank assembly and stability to synapses. J Neurosci 24, 9391-9404.  Rubenstein, J. L., and Merzenich, M. M. (2003). Model of autism: increased ratio of excitation/inhibition in key neural systems. Genes Brain Behav 2, 255-267.  Sabo, S. L., Gomes, R. A., and McAllister, A. K. (2006). Formation of Presynaptic Terminals at Predefined Sites along Axons. J Neurosci 26, 10813-10825.  Sabo, S. L., and McAllister, A. K. (2003). Mobility and cycling of synaptic protein-containing vesicles in axonal growth cone filopodia. Nat Neurosci 6, 1264-1269.  Sara, Y., Biederer, T., Atasoy, D., Chubykin, A., Mozhayeva, M. G., Sudhof, T. C., and Kavalali, E. T. (2005). Selective capability of SynCAM and neuroligin for functional synapse assembly. J Neurosci 25, 260-270.  Scheiffele, P., Fan, J., Choih, J., Fetter, R., and Serafini, T. (2000). Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell 101, 657-669.  Serge, A., Fourgeaud, L., Hemar, A., and Choquet, D. (2003). Active surface transport of metabotropic glutamate receptors through binding to microtubules and actin flow. J Cell Sci 116, 5015-5022.  Shapira, M., Zhai, R. G., Dresbach, T., Bresler, T., Torres, V. I., Gundelfinger, E. D., Ziv, N. E., and Garner, C. C. (2003). Unitary assembly of presynaptic active zones from Piccolo-Bassoon transport vesicles. Neuron 38, 237-252.  Sheetz, M. P., Steuer, E. R., and Schroer, T. A. (1989). The mechanism and regulation of fast axonal transport. Trends Neurosci 12, 474-478.  Song, J. Y., Ichtchenko, K., Sudhof, T. C., and Brose, N. (1999). Neuroligin 1 is a postsynaptic cell- adhesion molecule of excitatory synapses. Proc Natl Acad Sci U S A 96, 1100-1105.  Sumita, K., Sato, Y., Iida, J., Kawata, A., Hamano, M., Hirabayashi, S., Ohno, K., Peles, E., and Hata, Y. (2007). Synaptic scaffolding molecule (S-SCAM) membrane-associated guanylate kinase with inverted organization (MAGI)-2 is associated with cell adhesion molecules at inhibitory synapses in rat hippocampal neurons. Journal of Neurochemistry 100, 154-166.   148 Tai, C. Y., Mysore, S. P., Chiu, C., and Schuman, E. M. (2007). Activity-regulated N-cadherin endocytosis. Neuron 54, 771-785.  Talebizadeh, Z., Bittel, D. C., Veatch, O. J., Butler, M. G., Takahashi, T. N., and Miles, J. H. (2004). Do known mutations in neuroligin genes (NLGN3 and NLGN4) cause autism? J Autism Dev Disord 34, 735- 736.  Thompson, C., Lin, C. H., and Forscher, P. (1996). An Aplysia cell adhesion molecule associated with site- directed actin filament assembly in neuronal growth cones. J Cell Sci 109 ( Pt 12), 2843-2854.  Triller, A., and Choquet, D. (2005). Surface trafficking of receptors between synaptic and extrasynaptic membranes: and yet they do move! Trends Neurosci 28, 133-139.  Varoqueaux, F., Jamain, S., and Brose, N. (2004). Neuroligin 2 is exclusively localized to inhibitory synapses. Eur J Cell Biol 83, 449-456.  Waites, C. L., Craig, A. M., and Garner, C. C. (2005). MECHANISMS OF VERTEBRATE SYNAPTOGENESIS. Annual Review of Neuroscience 28, 251-274.  Washbourne, P., Bennett, J. E., and McAllister, A. K. (2002). Rapid recruitment of NMDA receptor transport packets to nascent synapses. Nat Neurosci 5, 751-759.  Washbourne, P., Liu, X. B., Jones, E. G., and McAllister, A. K. (2004). Cycling of NMDA receptors during trafficking in neurons before synapse formation. J Neurosci 24, 8253-8264.  Yamada, M., Hashimoto, T., Hayashi, N., Higuchi, M., Murakami, A., Nakashima, T., Maekawa, S., and Miyata, S. (2007). Synaptic adhesion molecule OBCAM; synaptogenesis and dynamic internalization. Brain Research 1165, 5-14.  Yan, J., Oliveira, G., Coutinho, A., Yang, C., Feng, J., Katz, C., Sram, J., Bockholt, A., Jones, I. R., Craddock, N., et al. (2004). Analysis of the neuroligin 3 and 4 genes in autism and other neuropsychiatric patients. Mol Psychiatry.  Ylisaukko-oja, T., Rehnstrom, K., Auranen, M., Vanhala, R., Alen, R., Kempas, E., Ellonen, P., Turunen, J. A., Makkonen, I., Riikonen, R., et al. (2005). Analysis of four neuroligin genes as candidates for autism.  Yu, W., Jiang, M., Miralles, C. P., Li, R.-w., Chen, G., and de Blas, A. L. (2007). Gephyrin clustering is required for the stability of GABAergic synapses. Molecular and Cellular Neuroscience 36, 484-500.  Zhang, W., and Benson, D. L. (2001). Stages of Synapse Development Defined by Dependence on F-Actin. J Neurosci 21, 5169-5181.  Ziv, N. E., and Garner, C. C. (2004). Cellular and molecular mechanisms of presynaptic assembly. Nat Rev Neurosci 5, 385-399.          149 4 Overall discussion  150 4.1 SUMMARY OF FINDINGS  The objective of this work was to determine:  (1) How pre-assembled postsynaptic scaffolding complexes participate in glutamatergic synapse development. (2) The molecular mechanisms that influence neuroligin trafficking within the neuronal membrane  Previous studies have documented non-synaptic postsynaptic protein complexes, and their function was unknown (Blue and Parnavelas, 1983; Fiala et al., 1998; Rao et al., 1998; Sans et al., 2003; Washbourne et al., 2002). Here we assessed trafficking of several scaffolding proteins in young neurons, prior to the majority of synapse formation. We found that: (1) stationary complexes containing PSD-95, GKAP and Shank are abundant in young neurons at both synaptic and non-synaptic sites, (2) a small fraction of this scaffold complex is mobile, and can be recruited to nascent and existing presynaptic contacts, (3) a subset of stationary protein complex clusters contains NL1 and recruits synaptophysin-positive axonal transport vesicles that can recycle the vital dye FM 4-64, (4) assembly of the scaffold protein complex requires PSD-95, as interfering with PSD- 95 expression by siRNA disrupts clustering of GKAP and Shank, and reduces the number of excitatory synapses in young hippocampal neurons, and finally (5) knock down of PSD-95 results in an overall increase in VGAT puncta positive for NL1, indicating a shift in NL1 localization from excitatory to inhibitory contacts. Endogenous NL1 is weakly clustered in young neurons and more enriched at synapses at later stages (Gerrow, 2005; Levinson et al., 2005), suggesting that adhesion complexes may be assembled at synapses through diffusion within the membrane, and trapping at the appropriate site. To test this hypotheis, we demonstrate the fluorescence recovery after photo-beaching (FRAP) dynamics of NL1 and NL2, and show that the mobile fraction of NL1 is larger compared to NL2. Interestingly, FRAP dynamics of AP-tagged NL clusters labeled by BirA, a method that specifically labels surface proteins, are  151 similar to GFP-tagged proteins, suggesting that most mobile neuroligin comes from a surface pool. Time-lapse of surface labeled NLs showed that NL1 is more mobile than NL2, and that c-terminal domains are important for the relative stability of NL2 clusters. The influence of c-terminal domains on the surface mobility of NLs was also measured using FRAP where deletion of the PDZ-binding motif of NL1 significantly increased its mobile fraction, and deletion of this motif in NL2 had no effect. In order to determine some of the molecular interactions influencing NL surface trafficking, we performed experiments in the presence of specific scaffold proteins. Exogenous expression of PSD- 95 compared to Shank-1b, increased the recovery time and reduced the mobile pool of NL1. S-SCAM, a scaffold at excitatory and inhibitory synapses, can cluster both NLs when over-expressed, and decrease the size of the mobile pools of both NLs. The importance of this decrease in mobile fraction can be measured by the increased presynaptic enhancement when neurons are transfected with S-SCAM and NL, when compared to NL alone. The actin and microtubule cytoskeleton also had differential effects on the mobility of neuroligins. Depolymerization of microtubules moderately decreased the mobile pool of NL1, whereas blocking the de-polymerization of the actin cytoskeleton dramatically decreased NL1 recovery.  Depolymerization of either actin or microtubules had a similar effect on NL2. Together these results demonstrate that pre-assembled scaffolding complexes are involved in the formation of glutamatergic synapses in early development, and dissect some of the molecular interactions important in their formation and function. These results also demonstrate that the trafficking of neuroligins occurs through lateral diffusion, and describe some of the intracellular interactions that can decrease this diffusion, effectively trapping NL at the appropriate site. By dissecting these events during synaptogenesis in young hippocampal neurons in vitro, this will hopefully bring a better understanding of mechanisms underlying synapse formation in vivo.  152 4.2 ROLE OF PRE-ASSEMBLED SCAFFOLDING COMPLEXES IN SYNAPSE FORMATION  4.2.1 How do pre-assembled complexes fit into out current knowledge of glutamatergic synapse formation?  On the presynaptic side, mechanisms that determine where along axons contacts are stabilized and synapses formed are not fully understood. Two scenarios can be imagined: (1) The entire axon could be equally competent for synaptogenesis, and synapse formation could occur wherever dendrites contact axons; or (2) synaptogenic interactions could be stabilized only at restricted locations along axons. In parallel, on the postsynaptic side the mechanisms that determine where along axons contacts are stabilized and synapses formed are not fully understood, and again, two scenarios may be envisaged: (1) The entire dendrites could be competent for synaptogenesis, and synapse formation could occur wherever axons contact dendrites; or (2) synaptogenic interactions could be stabilized only at restricted locations along dendrites. Current data suggest that pre-defined sites and contact mediated sites for synapse formation are important in the making of glutamatergic synapses (to be discussed more below). This concept that a synapse can be made in multiple ways is intriguing. Proper synapse formation is essential to the formation and function of the brain, without which, often an organism cannot survive. This it is not surprising that many mechanisms exist for the genesis of such an integral part of a living being. However, with so many mechanisms, the following questions remain: When is each mechanism employed? Are there common regulatory mechanisms? If you stop one method of synapse formation, how is this compensated? What about other synapses? Synapse formation it appears, is still ripe with questions for future research.  Initial formation of presynaptic terminals can occur preferentially at predefined sites within the axons of cortical neurons. Stable STV accumulation and formation of presynaptic terminals induced by neuroligin-expressing non-neuronal cells was restricted to particular sites within the axon. Although it is unclear which molecules comprise these  153 sites, STVs paused repeatedly at these predefined sites, even in the absence of contact with dendrites or glia (Sabo et al., 2006). Interestingly, the density of pause sites correlates well with the density of synapses observed in these neurons at 8±10 DIV ( 0.17±0.19 synapses/µm). Notably, recent reports using non-mammalian systems, including studies of synaptogenesis in zebrafish Mauthner cell  (Jontes et al., 2000) and Caenorhabditis elegans HSNL neuron (Shen and Bargmann, 2003), have demonstrated the utilization of predefined, reproducible sites for presynaptic terminal formation in vivo. Although the initial complement of en passant synapses appears to form at STV pause sites in young neurons, this mechanism cannot completely account for the high density of synapses in the adult brain. Synaptogenesis can also be initiated by axonal growth cones which may use separate mechanisms (Ahmari et al., 2000; Washbourne et al., 2002). Presynaptic proteins might be recruited to sites of preexisting clusters of postsynaptic scaffolding protein complexes which are abundant in young hippocampal neurons in vitro (Chapter 2; Rao et al., 1998; Gerrow et al., 2006) and have been observed by EM in vivo (Blue and Parnavelas, 1983; Fiala et al., 1998; Hinds and Hinds, 1976; Steward and Falk, 1986). However, it is unclear whether these sites of postsynaptic protein accumulation occurred at contacts with predefined axonal STV pause sites. In older neurons (DIV 14 or greater), glutamatergic synapses between axon and dendrite shafts of hippocampal neurons can form in about an hour after the initial accumulation of presynaptic vesicles. Presynaptic proteins, including synaptic vesicle precursors (STVs) and piccolo-transport vesicles (PTVs), are mobile in axons before synapses are formed. These precursors are the first proteins recruited to nascent synapses and after  approximately 30 minutes, PSD- 95 accumulates at these sites followed by glutamate receptors (Bresler et al., 2004; Friedman et al., 2000). Finally, at later stages of synaptogenesis (at approximately 16 DIV), new presynaptic boutons can be formed as a result of an existing functional bouton splitting into multiple new terminals (Krueger et al., 2003). Together these results suggest that multiple hierarchies can result in the formation of a new synapse, and that different strategies are more prevalent at different developmental stages.  We were able to capture both pre- and postsynaptic mechanisms of recruitment of synaptic proteins to nascent contacts in DIV 5-7 neurons (Chapter 2). There are inherent  154 advantages to both the presynaptic and postsynaptic modes of synapse formation for proper connectivity of a neuronal circuit. In early development, guidance cues drives an axon into its respective target field, contacts the correct neuron, and thus presynaptic mechanisms may ensure recruitment of the appropriate scaffold and receptors that matches the neurotransmitter present in axonal terminals.  Conversely, a dendrite primed with the appropriate postsynaptic complex may serve to determine the number of sites to be stabilized or eliminated upon encountering an axon en passant, and this may dictate the number and type of synapses a neuron receives. Considering that the preformed pool is scarce in older neurons, this postsynaptic mechanism may not significantly contribute to synapse formation in more mature neurons, and must rely on synapse addition by presynaptic induction on a point by point basis. Experiments studying the adhesion complex formed by NLs and NRXs suggest that both presynaptic and postsynaptic mechanisms stimulate synapse assembly: When expressed in non-neuronal cells, β-NRX is sufficient to drive the recruitment of postsynaptic proteins, and conversely NLs are sufficient to drive the recruitment of the presynaptic release machinery (Levinson et al., 2005). Studies suggest that scaffolding proteins play an important role in dictating the behavior of cell adhesion molecules in synapse development. For instance, PSD-95 enhances NL1 clustering and maturation of excitatory synapses at the expense of inhibitory contacts (Prange et al., 2004).  Thus, factors that govern early assembly of these proteins at non-synaptic sites may be critical for controlling the number of newly formed synapses.    155 4.2.2 What about other synapse types?  In addition to the glutamatergic pre-assembled complexes described herein (Chapter 2), other synapse types have been demonstrated to have pre-assembled clusters of proteins. The most infamous being the pre-patterning of acetylcholine receptors (AChRs) on the muscle fiber, independent of contact with motoneuron terminals (Arber, 2002; Lin, 2001; Yang, 2001), but dependent on the function of muscle-specific kinase (MuSK) (Lin, 2001; Yang, 2001). In vitro, synapse formation occurs when pre-patterning has not been observed, and in coming neurites do not preferentially contact pre-clustered AChR¶s (Anderson, 1977; Frank, 1979). However, in vivo AChR clusters are preferentially formed at the endplate band where innervation eventually occurs (Zhou, 1997). One may envision two scenarios: Motor axons ignore these preformed clusters in vivo just as they do in vitro, and use agrin to organize new clusters and the use a µsecond signal¶ to disperse non-synaptic clusters. Alternatively, axons might recognize these clusters or encounter them by chance, and then use agrin to enlarge and/or stabilize them. The latter appears to be the case based on work in zebrafish, where motor axon growth cones and filopodia are selectively extended toward and contact pre-patterned AChR clusters, followed by the rapid clustering of presynaptic vesicles and insertion of additional AChRs (Figure 4.1) (Panzer et al., 2006). This behavior may be similar to the growth cones that we observed (Chapter 2), and further investigation is warranted. In addition, recent evidence has suggested that this may also be true in mammals, and that the tyrosine kinase ErB2 may help establish the distribution of pre-patterned AChRs (Vock et al., 2008).   156  Figure 4.1: Axon filopodia preferentially extend toward and contact pre-patterned AChR clusters. A primary motor neuron, RoP, expressing VAMP-GFP (green) and pre-patterned AChR clusters labeled with rhodamine BTX (red) from a 24 hpf embryo. (A) Many filopodia are extended toward the caudal area containing pre-patterned AChR clusters (panel 0:00, arrow). Filopodia contact this AChR cluster twice (panels 0:10 and 3:30, asterisk) and retract. Scale bar, 5 µm. (B) The location of pre-patterned AChR clusters was aligned at 0ƒ (red dot). Each line represents one filopodia, the length of the line represents filopodia length, and the angle represents the angle of filopodia extension with respect to a pre-patterned AChR cluster. The majority of filopodia are extended at an angle 30ƒ with respect to pre-patterned AChR clusters (right panel). Distribution of the angle of filopodia extension with respect to a control area 15 µm away from the axon. Filopodia are extended randomly if an AChR cluster is not present (left panel). Adapted from (Panzer et al., 2006).  Interestingly, just like at the NMJ there is a measure of pre-patterning at inhibitory synapses: GABAA receptors form clusters on cell membranes even before synapse formation (Brunig et al., 2002; Christie et al., 2002a; Danglot et al., 2003; Studler et al., 2005), and are often smaller than those induced by GABAergic nerve terminals (Brunig et al., 2002; Christie et al., 2002a). Furthermore, the cell-specific expression and subcellular localization of GABAA receptor subtypes suggests that receptor targeting is unlikely to be determined by presynaptic innervation alone, and that protein±protein interaction mechanisms dependent on appropriate sequence motifs present in particular GABAA receptor subunits, are likely to ensure appropriate sorting and synaptic targeting.  157 Thus, the role of pre-assembled protein complexes and synaptic hot spots appears to be a feature of both presynaptic and postsynaptic terminals, and has been demonstrated in multiple synapse types, namely the NMJ, the glutamatergic synapse (Chapter 2), and the GABAergic synapse. Perhaps this concept may be further explored in yet more synapse types including those of the dopaminergic and serotonergic system. Research into µhot spots¶ for these types of synapses may yield intriguing results given the well-documented modulatory role these synapses have on other synapses. For instance, stimulation of D1 receptors results in the addition AMPA receptors at glutamatergic synapses (Gao et al., 2006; Gao and Wolf, 2007; Mangiavacchi and Wolf, 2004; Swayze et al., 2004). Thus, one may envisage this cross-talk between synapse types may be regulated by the spatial distribution of these synapses, and µhot spots¶ for dopaminergic input may be an important mechanism for regulating this spatial distribution. Although some of the molecular mechanisms for axon guidance that instruct domaminergic axons to establish highly stereotypic connections to the forebrain have been characterized (reviewed in Van den Heuvel and Pasterkamp, 2008), and  the structural and biochemical changes in domaminergic circuits have been associated with multiple psychiatric and neurodegenerative disorders, our understanding of the mechanisms that regulate the formation and maintenance of these circuits is rudimentary, and the signals that co- ordinate the formation of dopaminergic synapses are unknown.   4.2.3 The connection between synapse formation and neuron morphology  The dendritic arbor is responsible for receiving and consolidating neuronal input. Outgrowth and morphogenesis of the arbor are complex stages of development that are poorly understood. The space that a dendritic arbor occupies is determined largely by a combination of growth-promoting signals that regulate arbor size, chemotropic cues that steer dendrites into the appropriate space, and neurite-neurite contacts that ensure proper representation of the dendritic field and appropriate synaptic contacts  (reviewed in Cline  158 and Haas, 2008). Dendrite arbor development and synapse formation/maturation are concurrent (Figure 4.2 A), and provides the foundation of the synaptotrophic hypothesis of dendritic development. Evaluation of this hypothesis requires an understanding of the processes of synaptogenesis and synapse maturation as well as an accurate description of dendritic arbor development. Now that we have sufficiently detailed information describing these complex events from a number of systems including the retinotectal system of both Xenopus and Zebrafish, it has been possible to test whether synaptic inputs control morphological development (Cline and Haas, 2008).  For instance, the role of AM PA receptor (AMPAR)-mediated glutamatergic transmission in dendrite growth has been tested by expressing peptides corresponding to the intracellular C-terminal domains of AMPAR subunits GluR1 (GluR1Ct) and GluR2 (GluR2Ct) in optic tectal neurons of the Xenopus retinotectal system. These peptides significantly reduce AMPAR synaptic transmission in transfected neurons while leaving visual system circuitry intact. Daily in vivo imaging revealed that these peptides dramatically impaired dendrite growth, resulting in less complex arbors than controls (Figure 4.2 ) (Haas et al., 2006).  Figure 4.2:Dendritic arbor growth and synapse maturation are concurrent. (A) The diagram shows an immature neuron with a simple dendritic arbor and excitatory synapses which are predominated by NMDA-type glutamate receptors. As the neuron matures, the dendritic arbor becomes more complex and the synapses mature by adding AMPA-type glutamate receptors (adapted from (Cline and Haas, 2008)). (B)  In vivo time-lapse images acquired at 24-h intervals over 4 days starting 24 h after transfection. Control tectal neurons elaborate complex dendritic arbors by iterative extension and maintenance of interstitial branches from existing lower order branches. Tectal neurons expressing GluR1 or GluR2 C-terminal peptides develop simple arbors with fewer higher order branches. White arrows demarcate axons. Morphometric analysis of three-dimensional renderings from tectal measuring dendritic branch length (Left graph) and branch tip number (right graph). (adapted from (Haas et al., 2006)).  159 Recent findings have identified synaptic scaffolding proteins, in addition to their role in synapse formation, as regulators of dendritic morphology (Vessey and Karra, 2007). For instance, over-expression of PSD-95 in immature neurons led to dramatic alterations of the dendritic arbor. Primary dendrites were shorter, there were fewer secondary dendrites and overall dendritic complexity was reduced. Conversely, when PSD-95 protein levels were knocked down via antisense oligonucleotides, secondary dendrite numbers increased, resulting in a more complex dendritic arbor (Charych et al., 2006). Based on these findings, it appears that non-synaptic PSD-95, expressed in developing neurons, acts as a negative regulator of dendritic branching. Similarly, over-expression of Densin- 180, a member of the LRR and PDZ family of scaffolding proteins (Walikonis et al., 2001), developing neurons form irregular and substantially more complex dendritic arbors. These neurons also failed to develop synapses, as defined by dendritically localized PSD-95 clusters. Acting antagonistically, Shank proteins prevented the alterations in neuronal morphology induced by the over-expression of Densin-180 when co-expressed (Quitsch et al., 2005). Thus it appears that large macromolecular complexes containing scaffolding proteins, cytoskeletal regulatory proteins, and cytoskeletal motors are responsible for the effects found on dendritic patterning. Within this complex, the scaffolding proteins would appear to play multiple roles as adaptor molecules, regulators of effector proteins and as targeting elements for other proteins involved in cytoskeletal regulation. Although not limited to regulating dendritic pattern formation in the developing neuron, it would appear that this is the primary role of scaffolding proteins found outside of the mature synapse.   160 4.3 ROLE OF SURFACE MOBILITY AND INTRACELLULAR INTERACTIONS FOR ADHESION PROTEIN LOCALIZATION   4.3.1 How does this fit in with our current knowledge of adhesion protein trafficking?  Being a transmembrane protein, neuroligin trafficking could occur through direct insertion into postsynaptic membranes, or through extra-synaptic insertion followed by lateral diffusion. Such a hypothesis is similar to what has been proposed for neurotransmitter receptors at both glutamatergic and GABAergic synapses (Choquet and Triller, 2003; Triller and Choquet, 2005). Trafficking of the adhesion moelcule L1 in neuronal growth cones suggest L1 adhesions at growth cones form initially via L1 exocytosis and lateral diffusion, accompanied by a coupling to the actin flow (Dequidt et al., 2007). Similaily, it has been shown that the post-transltional addition of polysialic acid (PSA) onto the adhesion molecule NCAM, an important protein for modulating adhesion between cells and stimulating cell migration and neurite outgrowth, can influence its mobile fraction within the cell membrane (Conchonaud et al., 2007). Although these results suggest that lateral diffusion is likely to be important in the trafficking of other adhesion complexes, it is still unclear how these complexes are trafficked to synaptic sites.  Our results favour the trafficking of NLs to synapstic sites through lateral diffusion (Chapter 3). This result may explain why endoengous staining of NLs shows weak clusters with diffuse staining non-synaptically, and then appear more distinctly clustered at later developmental stages (Gerrow, 2005; Levinson et al., 2005). Other adhesion complexes have also shown this change in distribution with synapse maturation. For instance, in development, N-cadherin and β-catenin are distributed diffusely along the length of dendritic motile filopodia and upon contact with an axon, the cadherin complex accumulates at points of contact (Jontes et al., 2004; Togashi et al., 2002).   161 4.3.2 How does this fit in with our current knowledge of factors that affect neuroligin trafficking?   Mechanisms that govern differential trafficking and retention of adhesion proteins to synapses remain unclear. Furthermore, mechanisms that influence targeting to excitatory versus inhibitory are not fully understood. Because of the differential trafficking of neuroligin isoforms, it offers a great tool to study both of these processes.   4.3.2.1 Splice sites and neurexin binding   Splice alterations of NRXs and NLs have been demonstrated to contribute to differences in function at GABAergic versus glutamatergic synapses (Table 4.1). In co-culture assays, addition of the site 4 insert to ȕ-NRX reduces its ability to cluster the glutamate postsynaptic proteins NL1/3/4 and PSD-95, but not the GABA postsynaptic proteins NL2 and gephyrin (Chih et al., 2005; Graf et al., 2004). Consistent with this finding, and with the low affinity of +S4 ȕ-NRX for +B NL, addition of the B splice insert to NL1, or artificially to NL2, reduces their ability to cluster VGAT but not VGlut1 when over- expressed in neurons (Chih et al., 2005). Also consistent with this idea, NL2 (which is always íB) promotes VGAT clustering more than +B NL1 (Chih et al., 2005; Levinson et al., 2005). Thus, +S4 ȕ-NRX and íB NLs together selectively promote differentiation of GABAergic synapses, whereas íS4 ȕ-NRX and +B NL1 together selectively promote differentiation of glutamatergic synapses.  A few issues raised in these recent studies seem more controversial, including the extent to which splicing regulates localization of NLs. Chih et al. (2004) reported that recombinant +B NL1 is localized preferentially at glutamatergic synapses, whereas íB NL1 is localized equally at glutamatergic and GABAergic synapses. By contrast, Graf et al. (2005) reported that artificial addition of the B splice insert to NL2 did not alter exclusive localization of this protein to GABAergic synapses. These contrasting results with the B splice site may be reconciled when intracellular interactions are also  162 accounted. The artificial addition of the B splice site may not be able to significantly shift NL2 to excitatory synapses due to of intracellular domains and interactions with specific scaffolding proteins (Chapter 3). Thus, intracellular interactions may further potentiate to the preferential targeting of the NL isoforms by decreasing the mobility of NL isoforms at the correct site, may allow for the interaction between specific neuroligin and neurexin splice variants, further potentiating their retention at the appropriate site.   Table 4.1: Synapse selectivity of neurexins and neuroligins. Induction of postsynaptic protein clustering by +S4 neurexin 1ȕ in co-culture assaysa Proteins: Neuroligin 1/3/4 Neuroligin 2 PSD-95 Gephyrin Clustering (%): 65b 90 b 50 b, 7 c 104 b, 110 c Enhancement of selective presynaptic input onto neurons by over-expression of neuroligin variants d  Neuroligin 1 (+B) Neuroligin 1 (íB) Neuroligin 2 (always íB) Enhancement of VGlut1: 4.7 c, 1.5 e 2.5 c 5.3 c, 1.5 e Enhancement of VGAT: 1.5 c, 2.2 e 5.0 c 5.6 c, 3.0 e a Induction is normalized to 100% induction by neurexin 1ȕ (íS4). b (Graf et al., 2004) c (Chih et al., 2005)  d Enhancement is reported as increased density of inputs immunoreactive for VGlut (a marker of glutamatergic presynapses) or VGAT (a marker of GABAergic presynapses) relative to a value of 1.0 for control input density onto neurons expressing enhanced green fluorescent protein (EGFP). The neuroligin forms tested all contained the A-site insert. e (Levinson et al., 2005) Adapted with permission from (Craig and Kang, 2007).   4.3.2.2 Dendritic targeting Appropriate polarized targeting is particularly relevant for proteins that contribute to synapse formation: Contact of axons with dendrites initiates the development of synaptic structures whereas axo-axonal contacts generally do not lead to synapse assembly. This implies that there are directional synaptogenic signals provided by the dendrites that are absent from axons. Due to the heterophilic nature of the neuroligin/neurexin complex, adhesion of these proteins could provide a bidirectional signal for the assembly of synapses, and contribute to the structural asymmetry of neuronal synapses. A pre- requisite for such a role in synapse formation is that neuroligins should be exclusively  163 targeted to the somato-dendritic domain and should be excluded from axons. Indeed, exogenous expression of epitope-tagged NL1 proteins show that a 32-amino acid cytoplasmic sorting motif in NL1 is necessary and sufficient for dendritic targeting (Rosales et al., 2005). Conversely, one would predict that neurexins likewise should exclusively be targeted to axons, and excluded from dendrites. Interestingly, immuno-electron microscopy uncovered localization of neurexins to presynaptic terminals and active zones, but also uncovered an unexpected presence in the postsynaptic compartment. This cis-interaction of NRX-1ß with NL1 was shown to inhibit trans-binding, blocks the synaptogenic activity of NL1, and reduces the density of presynaptic terminals in cultured hippocampal neurons (Taniguchi et al., 2007). This demonstrates that the function of NRX proteins is more diverse than previously anticipated, and suggests that postsynaptic cis-interactions might provide a mechanism for silencing the activity of a synaptic adhesion complex. The pan-neurexin antibodies used for these ultra-structural studies recognized all neurexin isoforms, and therefore, it still remains to be shown whether a molecularly distinct subset of neurexin isoforms is predominantly enriched in dendrites, or whether many neurexin isoforms are represented in both the presynaptic and postsynaptic compartments.  4.3.2.3 Oligomerization  Oligomerization of NL1 is required for its function (Dean et al., 2003). Mutations of the amino acids K578/V579 and E584/L585 (located in an alpha helix at the base of the AChE-homologous domain important for dimerization) disrupt dimerization and subsequently, adhesion between NL and NRX-expressing cells. NL1 is primarily at glutamatergic and NL2 is primarily at GABAergic synapses in hippocampal neurons, and a fraction of each can be found at the 'wrong' synapse (Chih et al., 2005; Levinson et al., 2005). What about other neuroligin isoforms? Interestingly, NL3 can be found at both glutamatergic and GABAergic synapses (Budreck and Scheiffele, 2007).  Individual synapses contained both NL2 and -3, and co-immunoprecipitation studies revealed the presence of NL-1/3 and NL-2/3 complexes in brain extracts, suggesting that NL3 is a  164 shared component of glutamatergic and GABAergic synapses. Thus, in addition to extracellular and cytoplasmic interactions that can influence NL targeting, factors that influence NL dimerization may contribute to NL targeting.  4.3.3 Cytoplasmic interactions 4.3.3.1 Scaffolding proteins  Several scaffolding proteins have been demonstrated to interact with NL1,  NL2, and NL3 (see selected MAGUK examples in Table 4.2), including PDZ-domain scaffolding proteins such as PSD-95 and related MAGUKs, S-SCAM and related MAGIs, and probably also Shank, PICK1, GOPC and SPAR (Iida et al., 2004; Irie et al., 1997; Meyer et al., 2004).  However, the functional consequences for several of these interactions are poorly understood.  Table 4.2: Interactions between the cytoplasmic domains of neuroligins and various PDZ domains. Bait vector (in pBTM116)  Prey vector NL1-1 NL1-10 NL2-1 NL3 PSD-95 pVP16PSD95-2 (PDZ 1-3) 1437 ±  24 453 ± 10 2635 ± 48 1350 ±   6 pVP16SAP90-5 (PDZ 1) 46 ±   2 -- -- -- pVP16SAP90-6 (PDZ 2) 34 ±   0 -- -- -- pPrey514 (PDZ 3) 2068 ± 340 859 ± 32 466 ±  0 1809 ± 156 SAP102 pVP16SAP102-1 (PDZ 1-3) 885 ±   7 1046 ±  5 720 ± 13 789 ±   6 PSD-93 pVP16PSD93-1 (PDZ 1-3) 1394 ±  55 428 ± 18 596 ± 76 566 ±  10 DLG1 pVP16dlg-1 (PDZ 1-3) <20 -- <20 <20 Data list -galactosidase activities of yeast strains harboring the respective bait and prey plasmids. Single colonies from yeast cotransformed with the listed prey and bait vectors were selected on plates supplemented with minimal medium that lacked uracil, tryptophan, and leucine and were grown in liquid culture in the presence of selection medium for 40 hours. -Galactosidase activity and protein concentrations of cell extracts were determined in triplicate. Data shown are nanomoles of substrate hydrolyzed per minute per milligram of protein ± SD after background subtraction; --, not tested; <20, no detectable activity. Courtesy of (Irie et al., 1997).   165 To date, the best studied interaction is between NL1 and PSD-95. Over-expression of PSD-95 is able to restrict the action of NL1 to excitatory synapses (Levinson et al., 2005; Prange et al., 2004), whereas loss of PSD-95 can shift endogenous NL1 to inhibitory synapses (Chapter 2). Thus, it appears that the stoichiometry of adhesion to scaffold proteins is important in determining the trafficking/retention of NLs. We have extended these observations to also show that the scaffolding protein S-SCAM similarly can sequester the function of NL1 and NL2 (Chapter 3). The mechanisms by which NL2 is inhibited from binding to glutamate-specific PDZ domain proteins such as PSD-95 in neurons, how it localizes to GABAergic synapses and which proteins it binds to in neurons are questions under active investigation. Future studies into the precise function of the interaction with other scaffolding proteins, such as GRIP or PICK, may provide further insight into the mechanisms of NL targeting to excitatory versus inhibitory synapses.   4.3.3.2 Cytoskeleton  In addition to scaffolding molecules, the actin and microtubule cytoskeleton can structurally organize synapses.  Here, we described a link between the actin cytoskeleton and the movement of preassembled scaffolding complexes (Chapter 2), as well as a link between actin and microtubule cytoskeleton turn-over and the turn-over of NLs (Chapter 3). Our results with glutamatergic proteins are consistent with previous reports (Kuriu et al., 2006; Qualmann et al., 2004; Takahashi et al., 2003; Zhang and Benson, 2001), which together describe a strong co-relation between actin turn-over and the stability of many proteins at glutamatergic synapses. The interaction between the actin and microtubule cytoskeleton has also been reported for several proteins at inhibitory synapses (Bausen et al., 2006; Charrier et al., 2006; Fuhrmann et al., 2002; Giesemann et al., 2003; Kirsch and Betz, 1995; Ramming et al., 2000). Although this turn-over is less effected by the cytoskeleton in older neuronal cultures (Allison et al., 2000). Thus the cytoskeleton can affect the trafficking and retention of several classes of synaptic proteins, and offers a mechanism for determining the molecular content of both glutamatergic and GABAergic synapses. Because of the almost ubiquitous distribution of the cytoskeleton within  166 neurons, one may envisage mechanisms for local regulation, such as at specific dendritic spines, as well as homeostatic mechanisms that could affect the entire dendritic tree.   4.4 THERAPEUTIC TARGET  Signals received through excitatory versus inhibitory contacts control neuronal excitability and brain function. The importance of the development of proper synaptic balance is highlighted in when looking at instances where this is perturbed, as seen in mental diseases such as autism (Bartlett et al., 2005; Belmonte et al., 2004; Jamain et al., 2003; Laumonnier et al., 2004; Rubenstein and Merzenich, 2003; Talebizadeh et al., 2004; Yan et al., 2004; Ylisaukko-oja et al., 2005). Within this dissertation I  look at two aspects of synapse formation that may help in the establishment of the proper excitatory and inhibitory synaptic balance. First, I desribe a pre-assembled scaffolding complex which may help determine the number and position of glutamatergic synapses in hippocampal neurons during early development (Chapter 2). The loss of these pre-assembled complexes resulted in a shift of NL1 to inhibitory synapses. Second, I described the lateral trafficking of two adhesion molecules, NL1 and NL2, with distinct targeting to excitatory and inhibitory synapses, as well as some of the interations that influence this (Chapter 3). The imporance ot NLs in mental illness if futher supported by the presence of several mutations in neuroligin genes identified in autistic patients, and the altered social interaction, or µautistic behaviour¶ found in NL transgenic and knock-out mice (Chapter 1). Thus it appears that the trafficking of neuroligins is important for the establishment of synaptic balance, and provided with the knowledge of how to manipulate this trafficking, one may develop therapeutic strategies to manipulate neuroligin trafficking to re-establish proper synaptic balance.   167 4.5 CONCLUSION   Accumulating studies are leading to the realization that there might not be a single molecule, or molecular family, essential for assembly of CNS synapses. Nonetheless, neurexins and neuroligins are good candidates for central organizing molecules to stabilize networks of presynaptic and postsynaptic proteins across the synaptic cleft. Furthermore, stoichiometric interactions of neuroligins with select scaffolding proteins may help regulate this trans-synaptic function, and help in establishing the synaptic balance required for proper brain function. This thesis provides insight into some of the mechanisms that may aid in the establishment of synaptic balance. First, I describe the role of pre-assembled scaffold complexes in the establishment of excitatory synapses in early development. Secondly, I describe how neuroligins can diffuse on the neuron surface, and can be trapped at their appropriate sites through intracellular interactions. Further studies are needed to explore the trafficking of the rich diversity of neurexin and neuroligin variants; particularly useful will be targeted in vivo mutagenesis and analysis of the structure, molecular composition, function and long-term stability of glutamatergic and GABAergic synapses in defined circuits.  168 4.6 REFERENCES  Ahmari, S. E., Buchanan, J., and Smith, S. J. (2000). Assembly of presynaptic active zones from cytoplasmic transport packets. Nat Neurosci 3, 445-451.  Allison, D. W., Chervin, A. S., Gelfand, V. I., and Craig, A. M. (2000). Postsynaptic Scaffolds of Excitatory and Inhibitory Synapses in Hippocampal Neurons: Maintenance of Core Components Independent of Actin Filaments and Microtubules. J Neurosci 20, 4545-4554.  Anderson, M. J., Cohen, M.W. (1977). Nerve-induced and spontaneous redistribution of acetylcholine receptors on cultured muscle cells. J Physiol 268, 757-773.  Arber, S., Burden, S.J., Harris, A.J. (2002). Patterning of skeletal muscle. Curr Opin Neurobiol 12, 100- 103.  Bartlett, C. W., Gharani, N., Millonig, J. H., and Brzustowicz, L. M. (2005). Three autism candidate genes: a synthesis of human genetic analysis with other disciplines. Int J Dev Neurosci 23, 221-234.  Bausen, M., Fuhrmann, J. C., Betz, H., and O'Sullivan, G. A. (2006). The state of the actin cytoskeleton determines its association with gephyrin: Role of ena/VASP family members. Molecular and Cellular Neuroscience 31, 376-386.  Belmonte, M. K., Allen, G., Beckel-Mitchener, A., Boulanger, L. M., Carper, R. A., and Webb, S. J. (2004). Autism and abnormal development of brain connectivity. J Neurosci 24, 9228-9231.  Blue, M. E., and Parnavelas, J. G. (1983). The formation and maturation of synapses in the visual cortex of the rat. I. Qualitative analysis. J Neurocytol 12, 599-616.  Boucard, A. A., A.A., C., Comoletti, D., Taylor, P., and Sudhof, T. C. (2005). A splice code for trans- synaptic cell adhesion mediated by binding of neuroligin 1 to alpha- and beta-neurexins. Neuon 48, 229- 236.  Bresler, T., Shapira, M., Boeckers, T., Dresbach, T., Futter, M., Garner, C. C., Rosenblum, K., Gundelfinger, E. D., and Ziv, N. E. (2004). Postsynaptic density assembly is fundamentally different from presynaptic active zone assembly. J Neurosci 24, 1507-1520.  Budreck, E. C., and Scheiffele, P. (2007). Neuroligin-3 is a neuronal adhesion protein at GABAergic and glutamatergic synapses. European Journal of Neuroscience 26, 1738-1748.  Charrier, C., Ehrensperger, M.-V., Dahan, M., Levi, S., and Triller, A. (2006). Cytoskeleton Regulation of Glycine Receptor Number at Synapses and Diffusion in the Plasma Membrane. J Neurosci 26, 8502-8511.  Charych, E. I., Akum, B. F., Goldberg, J. S., Jornsten, R. J., Rongo, C., Zheng, J. Q., and Firestein, B. L. (2006). Activity-independent regulation of dendrite patterning by postsynaptic density protein PSD-95. J Neurosci 26, 10164-10176.  Chih, B., Engelman, H., and Scheiffele, P. (2005). Control of Excitatory and Inhibitory Synapse Formation by Neuroligins. Science 307, 1324-1328.  Choquet, D., and Triller, A. (2003). The role of receptor diffusion in the organization of the postsynaptic membrane. 4, 251-265.  Cline, H., and Haas, K. (2008). The regulation of dendritic arbor development and plasticity by glutamatergic synaptic input: a review of the synaptotrophic hypothesis. J Physiol 586, 1509-1517.  169  Conchonaud, F., Nicolas, S., Amoureux, M.-C., Menager, C., Marguet, D., Lenne, P.-F., Rougon, G., and Matarazzo, V. (2007). Polysialylation Increases Lateral Diffusion of Neural Cell Adhesion Molecule in the Cell Membrane. J Biol Chem 282, 26266-26274.  Craig, A. M., and Kang, Y. (2007). Neurexin-neuroligin signaling in synapse development. Current Opinion in Neurobiology Development 17, 43-52.  Dean, C., Scholl, F. G., Choih, J., DeMaria, S., Berger, J., Isacoff, E., and Scheiffele, P. (2003). Neurexin mediates the assembly of presynaptic terminals. Nat Neurosci 6, 708-716.  Dequidt, C., Danglot, L., Alberts, P., Galli, T., Choquet, D., and Thoumine, O. (2007). Fast Turnover of L1 Adhesions in Neuronal Growth Cones Involving Both Surface Diffusion and Exo/Endocytosis of L1 Molecules. Mol Biol Cell 18, 3131-3143.  Fiala, J. C., Feinberg, M., Popov, V., and Harris, K. M. (1998). Synaptogenesis via dendritic filopodia in developing hippocampal area CA1. J Neurosci 18, 8900-8911.  Frank, E., Fischbach, G.D. (1979). Early events in neuromuscular junction formation in vitro: induction of acetylcholine receptor clusters in the postsynaptic membrane and morphology of newly formed synapses. J Cell Biol 83, 143-158.  Friedman, H. V., Bresler, T., Garner, C. C., and Ziv, N. E. (2000). Assembly of new individual excitatory synapses: time course and temporal order of synaptic molecule recruitment. Neuron 27, 57-69.  Fuhrmann, J. C., Kins, S., Rostaing, P., El Far, O., Kirsch, J., Sheng, M., Triller, A., Betz, H., and Kneussel, M. (2002). Gephyrin interacts with Dynein light chains 1 and 2, components of motor protein complexes. J Neurosci 22, 5393-5402.  Gao, C., Sun, X., and Wolf, M. E. (2006). Activation of D1 dopamine receptors increases surface expression of AMPA receptors and facilitates their synaptic incorporation in cultured hippocampal neurons. J Neurochem 98, 1664-1677.  Gao, C., and Wolf, M. E. (2007). Dopamine alters AMPA receptor synaptic expression and subunit composition in dopamine neurons of the ventral tegmental area cultured with prefrontal cortex neurons. J Neurosci 27, 14275-14285.  Gerrow, K., Romorini, S., Nabi, S.,  Colicos, M., Sala, C.; El-Husseini, A. (2005). A preformed complex of postsynaptic proteins is involved in excitatory synapse development. Neuron 49, 547-562.  Giesemann, T., Schwarz, G., Nawrotzki, R., Berhorster, K., Rothkegel, M., Schluter, K., Schrader, N., Schindelin, H., Mendel, R. R., Kirsch, J., and Jockusch, B. M. (2003). Complex formation between the postsynaptic scaffolding protein gephyrin, profilin, and Mena: a possible link to the microfilament system. J Neurosci 23, 8330-8339.  Graf, E. R., Zhang, X., Jin, S. X., Linhoff, M. W., and Craig, A. M. (2004). Neurexins Induce Differentiation of GABA and Glutamate Postsynaptic Specializations via Neuroligins. Cell 119, 1013- 1026.  Haas, K., Li, J., and Cline, H. T. (2006). AMPA receptors regulate experience-dependent dendritic arbor growth in vivo. Proceedings of the National Academy of Sciences 103, 12127-12131.  Hinds, J. W., and Hinds, P. L. (1976). Synapse formation in the mouse olfactory bulb. II. Morphogenesis. J Comp Neurol 169, 41-61.   170 Iida, J., Hirabayashi, S., Sato, Y., and Hata, Y. (2004). Synaptic scaffolding molecule is involved in the synaptic clustering of neuroligin. Mol Cell Neurosci 27, 497-508.  Irie, M., Hata, Y., Takeuchi, M., Ichtchenko, K., Toyoda, A., Hirao, K., Takai, Y., Rosahl, T. W., and Sudhof, T. C. (1997). Binding of Neuroligins to PSD-95. Science 277, 1511-1515.  Jamain, S., Quach, H., Betancur, C., Rastam, M., Colineaux, C., Gillberg, I. C., Soderstrom, H., Giros, B., Leboyer, M., Gillberg, C., et al. (2003). Mutations of the X-linked genes encoding neuroligins NLGN3 and NLGN4 are associated with autism. Nat Genet.  Jontes, J. D., Buchanan, J., and Smith, S. J. (2000). Growth cone and dendrite dynamics in zebrafish embryos: early events in synaptogenesis imaged in vivo. Nat Neurosci 3, 231-237.  Jontes, J. D., Emond, M. R., and Smith, S. J. ( 2004). In vivo trafficking and targeting of N-cadherin to nascent presynaptic terminals. J Neurosci 24, 9027-9034.  Khosravani, H., Altier, C., Zamponi, G. W ., and Colicos, M. A. (2005). The Arg473Cys-neuroligin-1 mutation modulates NMDA mediated synaptic transmission and receptor distribution in hippocampal neurons. FEBS Letters 579, 6587-6594.  Kirsch, J., and Betz, H. (1995). The postsynaptic localization of the glycine receptor-associated protein gephyrin is regulated by the cytoskeleton. J Neurosci 15, 4148-4156.  Krueger, S. R., Kolar, A., and Fitzsimonds, R. M. (2003). The Presynaptic Release Apparatus Is Functional in the Absence of Dendritic Contact and Highly Mobile within Isolated Axons. Neuron 40, 945-957.  Kuriu, T., Inoue, A., Bito, H., Sobue, K., and Okabe, S. (2006). Differential Control of Postsynaptic Density Scaffolds via Actin-Dependent and -Independent Mechanisms 10.1523/JNEUROSCI.0522-06.2006. J Neurosci 26, 7693-7706.  Laumonnier, F., Bonnet-Brilhault, F., Gomot, M., Blanc, R., David, A., Moizard, M. P., Raynaud, M., Ronce, N., Lemonnier, E., Calvas, P., et al. (2004). X-Linked Mental Retardation and Autism Are Associated with a Mutation in the NLGN4 Gene, a Member of the Neuroligin Family. Am J Hum Genet 74, 552-557.  Levinson, J. N., Chery, N., Huang, K., Wong, T. P., Gerrow, K., Kang, R., Prange, O., Wang, Y. T., and El-Husseini, A. (2005). Neuroligins Mediate Excitatory and Inhibitory Synapse Formation: INVOLVEMENT OF PSD-95 AND NEUREXIN-1{beta} IN NEUROLIGIN-INDUCED SYNAPTIC SPECIFICITY. J Biol Chem 280, 17312-17319.  Lin, W., Burgess, R.W., Dominguez, B., Pfaff, S.L., Sanes, J.R., Lee, K.F. (2001). Distinct roles of nerve and muscle in postsynaptic differentiation of the neuromuscular synapse. Nature 410, 1057-1064.  Mangiavacchi, S., and Wolf, M. E. (2004). D1 dopamine receptor stimulation increases the rate of AMPA receptor insertion onto the surface of cultured nucleus accumbens neurons through a pathway dependent on protein kinase A. J Neurochem 88, 1261-1271.  Meyer, G., Varoqueaux, F., Neeb, A., Oschlies, M., and Brose, N. (2004). The complexity of PDZ domain- mediated interactions at glutamatergic synapses: a case study on neuroligin. Neuropharmacology 47, 724- 733.  Panzer, J. A., Song, Y., and Balice-Gordon, R. J. (2006). In Vivo Imaging of Preferential Motor Axon Outgrowth to and Synaptogenesis at Prepatterned Acetylcholine Receptor Clusters in Embryonic Zebrafish Skeletal Muscle. J Neurosci 26, 934-947.   171 Prange, O., Wong, T. P., Gerrow, K., Wang, Y. T., and El-Husseini, A. (2004). A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. PNAS 101, 13915-13920.  Qualmann, B., Boeckers, T. M., Jeromin, M., Gundelfinger, E. D., and Kessels, M. M. (2004). Linkage of the Actin Cytoskeleton to the Postsynaptic Density via Direct Interactions of Abp1 with the ProSAP/Shank Family. J Neurosci 24, 2481-2495.  Quitsch, A., Berhorster, K., Liew, C. W., Richter, D., and Kreienkamp, H. J. (2005). Postsynaptic shank antagonizes dendrite branching induced by the leucine-rich repeat protein Densin-180. J Neurosci 25, 479- 487.  Ramming, M., Kins, S., Werner, N., Hermann, A., Betz, H., and Kirsch, J. (2000). Diversity and phylogeny of gephyrin: tissue-specific splice variants, gene structure, and sequence similarities to molybdenum cofactor-synthesizing and cytoskeleton-associated proteins. Proc Natl Acad Sci U S A 97, 10266-10271.  Rao, A., Kim, E., Sheng, M., and Craig, A. M. (1998). Heterogeneity in the molecular composition of excitatory postsynaptic sites during development of hippocampal neurons in culture. J Neurosci 18, 1217- 1229.  Rosales, C. R., Osborne, K. D., Zuccarino, G. V., Scheiffele, P., and Silverman, M. A. (2005). A cytoplasmic motif targets neuroligin-1 exclusively to dendrites of cultured hippocampal neurons. European Journal of Neuroscience 22, 2381-2386.  Rubenstein, J. L., and Merzenich, M. M. (2003). Model of autism: increased ratio of excitation/inhibition in key neural systems. Genes Brain Behav 2, 255-267. Sabo, S. L., Gomes, R. A., and McAllister, A. K. (2006). Formation of Presynaptic Terminals at Predefined Sites along Axons. J Neurosci 26, 10813-10825.  Sans, N., Prybylowski, K., Petralia, R. S., Chang, K., Wang, Y. X., Racca, C., Vicini, S., and Wenthold, R. J. (2003). NMDA receptor trafficking through an interaction between PDZ proteins and the exocyst complex. Nat Cell Biol 5, 520-530.  Shen, K., and Bargmann, C. I. (2003). The immunoglobulin superfamily protein SYG-1 determines the location of specific synapses in C. elegans. Cell 112, 619-630.  Steward, O., and Falk, P. (1986). Protein-synthetic machinery at postsynaptic sites during synaptogenesis: a quantitative study of the association between polyribosomes and developing synapses. J Neurosci 6, 412- 423.  Swayze, R. D., Lise, M. F., Levinson, J. N., Phillips, A., and El-Husseini, A. (2004). Modulation of dopamine mediated phosphorylation of AMPA receptors by PSD-95 and AKAP79/150. Neuropharmacology 47, 764-778.  Takahashi, H., Sekino, Y., Tanaka, S., Mizui, T., Kishi, S., and Shirao, T. (2003). Drebrin-Dependent Actin Clustering in Dendritic Filopodia Governs Synaptic Targeting of Postsynaptic Density-95 and Dendritic Spine Morphogenesis. J Neurosci 23, 6586-6595.  Talebizadeh, Z., Bittel, D. C., Veatch, O. J., Butler, M. G., Takahashi, T. N., and Miles, J. H. (2004). Do known mutations in neuroligin genes (NLGN3 and NLGN4) cause autism? J Autism Dev Disord 34, 735- 736.  Taniguchi, H., Gollan, L., Scholl, F. G., Mahadomrongkul, V., Dobler, E., Limthong, N., Peck, M., Aoki, C., and Scheiffele, P. (2007). Silencing of Neuroligin Function by Postsynaptic Neurexins. J Neurosci 27, 2815-2824.   172 Togashi, H., Abe, K., Mizoguchi, A., Takaoka, K., Chisaka, O., and Takeichi, M. (2002). Cadherin regulates dendritic spine morphogenesis. Neuron 35, 77-89.  Triller, A., and Choquet, D. (2005). Surface trafficking of receptors between synaptic and extrasynaptic membranes: and yet they do move! Trends Neurosci 28, 133-139.  Van den Heuvel, D. M. A., and Pasterkamp, R. J. (2008). Getting connected in the dopamine system. Progress in Neurobiology 85, 75-93.  Vessey, J. P., and Karra, D. (2007). More than just synaptic building blocks: scaffolding proteins of the post-synaptic density regulate dendritic patterning. Journal of Neurochemistry 102, 324-332.  Vock, V. M., Ponomareva, O. N., and Rimer, M. (2008). Evidence for Muscle-Dependent Neuromuscular Synaptic Site Determination in Mammals. J Neurosci 28, 3123-3130.  Walikonis, R. S., Oguni, A., Khorosheva, E. M., Jeng, C. J., Asuncion, F. J., and Kennedy, M. B. (2001). Densin-180 forms a ternary complex with the (alpha)-subunit of Ca2+/calmodulin-dependent protein kinase II and (alpha)-actinin. J Neurosci 21, 423-433.  Washbourne, P., Bennett, J. E., and McAllister, A. K. (2002). Rapid recruitment of NMDA receptor transport packets to nascent synapses. Nat Neurosci 5, 751-759.  Yan, J., Oliveira, G., Coutinho, A., Yang, C., Feng, J., Katz, C., Sram, J., Bockholt, A., Jones, I. R., Craddock, N., et al. (2004). Analysis of the neuroligin 3 and 4 genes in autism and other neuropsychiatric patients. Mol Psychiatry.  Yang, X., Arber, S., William, C., Li, L., Tanabe, Y., Jessell, T.M., Birchmeier, C., Burden, S.J. (2001). Patterning of muscle acetylcholine receptor gene expression in the absence of motor innervation. Neuon 30, 399-410.  Ylisaukko-oja, T., Rehnstrom, K., Auranen, M., Vanhala, R., Alen, R., Kempas, E., Ellonen, P., Turunen, J. A., Makkonen, I., Riikonen, R., et al. (2005). Analysis of four neuroligin genes as candidates for autism.  Zhang, W., and Benson, D. L. (2001). Stages of Synapse Development Defined by Dependence on F-Actin. J Neurosci 21, 5169-5181.  Zhou, H., T Muramatsu, W Halfter, KW Tsim, HB Peng (1997). A role of midline in the development of the neuromuscular junction. Mol Cell Neurosci 10, 56-70.     173 5 Additional materials and methods  5.1 FILTER SETS Filter sets from Chroma include: CFP (D436/20x, 455CLP, D480/40m), GFP (HQ470/40x, Q494LP, HQ525/50m), YFP (HQ500/20x, Q515LP, HQ535/30m), DsRED (HQ535/50x, Q564LP, HQ610/75m), and FM 4-64 (HQ535/50x, Q564LP, HQ785/60m). No visible bleed-through or cross-excitation was detectable at higher exposures (3000ms) compared to those typically used for this study (300-800ms)(Figure 5.1).    Figure 5.1: Sample images for bleed-through and cross-excitation testing. Left panels display overlay, middle panels display images taken with appropriate filter set, and right panels show images taken with inappropriate filter set  (A) A neuron transfected with PSD-95 GFP with shows signal with the GFP filter, but not the DsRED filter. (B) A neuron transfected with CFP with shows signal with the CFP filter, but not the YFP filter. (C) A neuron transfected with PSD-95 YFP with shows signal with the YFP filter, but not the CFP filter.   174 5.2 IMAGE ANALYSIS  Images were processed at a constant threshold level (of 32,000 pixel values) to create a binary (µµmask¶¶) image, which was multiplied with the original image by using Boolean image arithmetic¶s. The resulting image contained a discrete number of clusters with pixel values of the original image. Only clusters with average pixel values 1.5 times greater than background pixel values were used for analysis. For co-localization analysis, background-subtracted immunofluorescence values for all imaging channels (red, green, and blue) were correlated within each dendrite mask. Co-localization was scored if clusters in two color channels were overlapping by at least 5 pixels for two postsynaptic clusters, and at least 1 pixel for a postsynaptic and presynaptic protein. Movement of clusters was analyzed using Image J (Wayne Rasband, NIH). Images were corrected for drift (RegisterROI, Michael Abràmoff, University of Iowa Hospitals and Clinics, Iowa), and velocities recorded (Manual Tracker, Fabrice Cordelières, Institut Curie, Orsay). Changes in position that were less than 2 pixel values per time point were omitted.  Confocal images for siRNA experiments were acquired with a Bio-Rad (Hercules, CA) MRC1024 microscope, using a Nikon (Tokyo, Japan) 60x objective with a sequential- acquisition setting at a resolution of 1280 x 1024 pixels. Each image was a z-series projection taken at 0.75 µm intervals. The morphometric measurements were made using MetaMorph image analysis software (Universal Imaging, West Chester, PA). For triple co-localization, the site of co-localization of the first two proteins is assessed for the presence of the third protein of interest. Triple co-localization was then expressed in terms of the percentage of co-clusters of the first two proteins assessed.  The distance traveled by these proteins ranged from 5 to 15 m. Mean velocity was calculated by measuring unidirectional movement that were at least 2 micron in distance during the imaging process  (1 to 2.5 hours period). Clusters were defined as stable if they remain stationary throughout the imaging period, ranging from 15 min, up to 120 min. Average velocity did not significantly change with inter-frame intervals, 0.80±0.12 µm/min at 2 min inter-frame interval versus 0.86±0.13 µm/min at 5 min inter-frame  175 interval, p=0.5, n=8,5 neurons; 82, 39 puncta. The proportion of moving clusters did not change with interframe interval, 9.6±1.8% at 2 min inter-frame interval versus 11.6±1.8% at 5 min inter-frame interval, p=0.3.   5.3 ANALYSIS OF EXPRESSION OF TAGGED PROTEINS  Various fluorescently tagged postsynaptic proteins were analyzed within 36 hours to minimize protein over-expression. Within this period, both endogenous and exogenous PSD-95, GKAP and Shank faithfully co-localized (87.6±2.8% endogenous versus 103.9±4.5% exogenous), and were found at both synaptic and non-synaptic sites as assessed by synaptophysin staining (47.2±5.2%  endogenous versus 50.7±10.9% exogenous, n=12 neurons per group, •600 puncta; Figure 5.2 A,B). Expression levels of PSD-95 GFP, GKAP DsRED and Shank CFP was assessed by immunostaining, and was compared to levels in untransfected neurons in the same field of view (Figure 5.2 C,D,E). Distribution of the fluorescent fusion proteins and immuno-reactivity completely overlapped in transfected neurons, demonstrating that GFP fluorescence was representative of the entire pool in these neurons. PSD-95 GFP expression was 117±7% for PSD-95 GFP, 116±17% for GKAP DsRED, and 126±13% for Shank CFP (n=12 neurons per group, p>0.6); far less than the 5-10 times greater expression level reported to significantly alter functional and structural characteristics of glutamatergic synapses (El-Husseini et al., 2000); and similar to other studies that showed that brief expression of synaptic proteins does not result in aberrant protein trafficking or enhanced protein function (Bresler et al., 2001; Bresler et al., 2004; Friedman et al., 2000; Okabe et al., 2001; Sabo and McAllister, 2003; Shapira et al., 2003; Washbourne et al., 2002). Cluster per length was not significantly different for PSD-95 GFP (107±17%), GKAP DsRED (95±8%), or Shank CFP (132±13, n=12 neurons per group, p>0.6, Figure 5.2E).    176   Figure 5.2: Characterization of expression of PSD-95 GFP, GKAP DsRED, and Shank CFP. (A) PSD-95 GFP and GKAP DsRED were transfected into DIV 5 neurons and expression was assessed for co-localization with VGLUT at DIV 6. Open arrows demarcate non-synaptic co-clusters of PSD-95 GFP and GKAP DsRED. (B) Quantitative analysis of the localization of exogenous PSD-95 GFP and GKAP DsRED (white bars) versus endogenous PSD-95 and GKAP (black bars). Exogenous and endogenous proteins had similar co-localization with each other (left pair of bars), and with presynaptic terminals as assessed by VGLUT staining (right pair of bars, p>0.8, n=12 neurons per group, >600 puncta). (C) Representative image of the level of expression in untransfected neurons (left panels) versus PSD-95 GFP transfected neurons (right panels). (D) The level of expression of fluorescent fusion constructs relative to endogenous expression in untransfected neurons was 117±7%  for PSD-95 GFP, 116±17% for GKAP DsRED, and 126±13% for Shank CFP (p>0.6). (E) The number of puncta per 10µm in neurons transfected with fluorescent fusion constructs relative to the endogenous number of puncta per 10µm was 107±17 for PSD-95 GFP, 95±8% for GKAP DsRED, and 132±13% for Shank CFP (n=12 neurons per group, p>0.6). Scale bars, (A,C) 1µm.   Level of expression of NLG1 CFP was 120±4% (p=0.2), and puncta per length was 88±15% (p=0.5) untransfected controls (n=10 neurons). VGLUT intensity of untransfected (139±10 AU), was not significantly increased in neurons transfected with NLG1 CFP alone (192±2 AU, p=0.07, n=10 neurons), versus NLG1 CFP and PSD-95 YFP (190±2 AU, p=0.07, n=10 neurons). NLG1 CFP clustering was not enhanced at sites with PSD-95 YFP (2.1±0.5 AU) versus clusters in neurons transfected with NLG1 CFP alone (1.7±0.2 AU, p=0.5, n=10 neurons).  Proportion of stationary co-clusters of PSD- 95 YFP when co-transfected with NLG1 CFP (87.8±2.5%) compared to with GKAP CFP (77.9±2.2%), and mobile co-clusters of PSD-95 YFP when co-transfected with NLG1  177 CFP (8.2±2.2%) compared to with GKAP CFP (10.5±0.9%, n=10 neurons each, p=0.3). Co-localization of preformed complex was not different between endogenous (56.5±3.6%, n=11 neurons) and exogenous (57.2±5.2%, n=10 neurons, p=0.7). Colocalization of NLG1 and VGLUT was not different between endogenous (28.1±5.3%, n=11 neurons) and exogenous (38.7±4.7%, p=0.7, n=10 neurons).    Figure 5.3: Characterization of exogenous expression of PSD-95 GFP, and neuroligin CFP. (A) Neurons transfected with NLG1 CFP alone or (B) with PSD-95 YFP. (C)  Level of expression of NLG1 CFP was 120±4% (p=0.2), and puncta per length was 88±15% (p=0.5) untransfected controls (n=10 neurons). (D) VGLUT intensity of untransfected cells, was not significantly increased in neurons transfected with NLG1 CFP alone (p=0.07, n=10 neurons), versus NLG1 CFP and PSD-95 YFP (p=0.07, n=10 neurons). (E) NLG1 CFP clustering was not enhanced at sites with PSD-95 YFP, versus clusters in neurons transfected with NLG1 CFP alone (p=0.5, n=10 neurons).  (F) Proportion of stationary clusters of PSD-95 YFP was not changed when co-expressed with NLG1 CFP (87.8±2.5%) or with GKAP CFP (77.9±2.2%).  Similarly, proportion of mobile clusters of PSD-95 YFP was not changed when co-expressed with NLG1 CFP (8.2±2.2%) or with GKAP CFP (10.5±0.9%, n=10 neurons each, p=0.3) (G) Co- localization of preformed complexes was not different between endogenous (56.5±3.6%, n=11 neurons) and exogenous (57.2±5.2%, n=10 neurons, p=0.7). (H) Co-localization of NLG1 and VGLUT was not significantly different between endogenous (28.1±5.3%, n=11 neurons) and exogenous (38.7±4.7%, p=0.7, n=10 neurons). Scale bars, (A,B) 5µm.    178 Expression of SYN DsRED by transfection with electroporation did not result in a significant protein over-expression (p=0.3), increased cluster area (p=0.6), nor enhanced PSD-95 clusters at sites of accumulation (p=0.5; Figure 5.4).   Figure 5.4: Expression of DsRED synaptophysin. Expression of SYN DsRED by transfection with electroporation at DIV0, assessment was at DIV7. This did not result in a significant protein over-expression (p=0.3), increased cluster area (p=0.6), nor enhanced PSD-95 clusters at sites of accumulation (p=0.5). Error bars are S.E.M. where n=10, and represents the number of fields of view.  In order to determine the mobile fraction of NL isoforms at their preferred sites, we preformed fluorescence recovery after photo-beaching (FRAP) on AP-tagged NL1 at Shank sites, and NL2 at gephyrin sites. As previously described, transfections were performed 24-36 hours prior to imaging using conditions that minimize over-expression and ensure targeting similar to endogenous proteins (Figure 5.5, 5.6). The average intensity of NL in APNL1 transfected neurons was 111.7±3.4% compared to controls (n=18). The average intensity of NL in APNL2 transfected neurons is 145.3±9.7% compared to controls (n=8). The average intensity of SYN clusters apposing APNL1 transfected neurons was 132.2±0.1% compared to controls (n=14). The average intensity of SYN clusters apposing APNL2 transfected neurons was 113.2±0.1% compared to controls (n=9). The recovery of AP-tagged NL1 and NL2 did not significantly differ with co-expression with GFP, GFPNL1, GFPNL2, or with appropriate scaffolds after 2 days expression (Figure 5.7).   179  Figure 5.5: AP-tagged NL expression analysis. (A) Representative images of neurons expressing APNL1 and stained with pan ĮNL antibody (left panels) or AP-NL2 and stained with ĮNL2 (right panels). (B) Quantification and normalization to un-transfected neurons within the same field of view show that the average intensity of NL in APNL1 transfected neurons was 111.7±3.4% compared to controls (n=18). The average intensity of NL in APNL2 transfected neurons is 145.3±9.7% compared to controls (n=8). (C) Representative images of neurons expressing APNL1 and stained (left panels) or AP-NL2 (right panels) and stained with ĮSYN. (D) The average intensity of SYN clusters apposing APNL1 transfected neurons was 132.2±0.1% compared to controls (n=14). The average intensity of SYN clusters apposing APNL2 transfected neurons was 113.2±0.1% compared to controls (n=9). Scale barsѽҏ5µm. Error bars indicate S.E.M., and n represents the number of neurons analyzed from 2 independent experiments.      Figure 5.6: AP-tagged NL distribution. (A) Example of the localization of most AP-NL1 clusters with YFP-Shank. High magnification of the box in the left panels is displayed on the right. (B) Example of the localization AP-NL2 clusters with CFP- gephyrin. High magnification of the box in the left panels is displayed on the right. Scale bars: 5µmѽҏңµm .     180   Figure 5.7: AP-tagged NL FRAP with GFP.  (A) FRAP curves comparing the recovery of AP NL1 when expressed with GFP, GFP NL1, GFP NL2 or YFP-Shank. The mobile fraction of AP NL1 with GFP was  67.7±8.2%, and was not significantly different when expressed with GFP NL1 (51.0±6.5%, p=0.04), GFP NL2 (54.6±5.9% , p=0.1), or YFP Shank (70.5±3.6%, p=0.7). (B) FRAP curves comparing the recovery of AP NL2 when expressed with GFP, GFP NL1, GFP NL2 or CFP-gephyrin. The mobile fraction of AP NL2 with GFP was  37.4±3.4%, and was 53.1±5.6%, (p=0.01) when expressed with GFP NL1, 58.6±11.8%, (p=0.07) when expressed with GFP NL2, and 48.5±3.1%, (p=0.04) when expressed with CFP gephyrin. Error bars indicate S.E.M., and n represents the number of clusters analyzed from 8-15 neurons, 2-4 independent experiments.  181 5.4 EVALUATION OF FM 4-64  15 µM FM 4-64 (Molecular Probes) was loaded for 30 seconds into presynaptic terminals using a hyperkalemic solution of 90mM KCl2 in modified HBSS, where equimolar NaCl2 was omitted for final osmolality of 310 mOsm. Neurons were rinsed three times and maintained in HBSS without Ca2+ and in the presence of 5 µM Mg2+ to prevent unloading during image acquisition. 1mM ADVESAP-7 (Sigma) was added to quench non-specific signal. A minimum of three images were captured to confirm positive sites FM loading were stationary presynaptic terminals and not orphan sites (Krueger et al., 2003). Unloading was performed for 30 seconds in the same hyperkalemic solution, and washed three times with NeuroBasal media for continued imaging. Unloading of FM  4-64, was observed in most of labeled presynaptic terminals, and puncta that did not significantly unload to 15% of initial intensity were not included in analysis. To confirm the fidelity of this dye to label excitatory presynaptic terminals, sites positive for FM 4-64 (Appendix Figure 5.8 A) were also labeled for the excitatory presynaptic marker VGLUT (83.6±19%, n=9 neurons, 367 puncta; Appendix Figure 5.8 B).   Figure 5.8: Fidelity of FM 4-64. The ability of FM 4-64 to label excitatory presynaptic terminals was determined in PSD-95 YFP transfected neurons by loading (A) and unloading of FM4-64, and subsequent retrospective immunostaining for VGLUT (B), 83.6±19%, n=9 cells, 367 puncta). Scale bar, 5µm.    182 5.5 SYNTHESIS OF MONOVALENT STREPTADVIDIN  Monovalent streptavidin was obtained from the lab of Dr. A.Y. Ting at the Massachusetts Institute of Technology (MIT), Cambridge. Streptavidin and avidin are used ubiquitously because of the remarkable affinity of their biotin binding, but they are tetramers, which disrupts many of their applications. Making either protein monomeric reduces affinity by at least 104-fold because part of the binding site comes from a neighboring subunit. The strategy for producing monovalent streptavidin by the Ting lab is shown in Figure 5.9A, where a streptavidin tetramer consisting of three subunits unable to bind biotin and one subunit that binds biotin as well as wild-type streptavidin is produced.  Howarth et al. (2005) found that a triple mutant (N23A, S27D, S45A; Figure 5.9 B) had negligible biotin binding, but left the tetramer structure intact. The biotin affinity of this mutant (composed of 'dead' (D) subunits) was so weak that it was difficult to measure, but we obtained an approximate Kd of 1.2 10-3 M. To generate monovalent streptavidin (Appendix Figure 6C), a 6xHis tag was added to the wild-type subunit ('alive' (A) subunit), and then D and A subunits were combined at a molar ratio of 3:1 in guanidinium hydrochloride, and then refolded rapidly by diluting the mixture into phosphate buffered saline (PBS). This refold generated a statistical mixture of tetramers of different composition. We purified the different tetramers using a nickel± nitrilotriacetic acid (Ni-NTA) column, eluting according to the number of 6His tags with increasing concentrations of imidazole. The tetramers could be distinguished by SDS- PAGE, if the samples were not boiled, according to the number of 6His tags present, showing that at least 30% of the initial mixture was of the monovalent A1D3 form (Appendix Figure 6D). Thus we obtained purified fractions of the monovalent A1D3 (final yield, 2 mg/l), as well as of the other chimeric streptavidins, A2D2 and A3D1. We confirmed the tetramer composition by boiling the samples before loading on SDS- PAGE, to determine the ratio of A to D subunits (Appendix Figure 6E), and by electrospray ionization mass spectrometry (found online www.nature.com/nmeth/journal/v3/n4/extref/nmeth861-S1.pdf). Despite the large mass of the streptavidin tetramer and the monovalent interaction between subunits, there is good agreement between expected and observed masses for D4, A1D3, A2D2, A3D1 and  183 A4. This engineered monovalent streptavidin is stable, since the subunit composition over time did not re-arrange when incubated at 26 ƒC or at 37 ƒC for up to one week. And importantly, has a binding affinity for biotin that is comparable to wild-type (A4 was 4.4 10-14 1.1 10-14 M (s.e.m.), similar to monovalent A1D3)(Howarth et al., 2006).    Figure 5.9: Generation of monovalent streptavidin. a) Wild-type streptavidin is a tetramer with four biotin binding sites (B, biotin). Monovalent streptavidin is a tetramer with 3 inactive subunits (dark gray) and one subunit that binds biotin with wild-type affinity (light gray). (b) Biotin binding site of wild-type streptavidin (from Protein Data Bank 1MK5), highlighting the three residues mutated to create the 'dead' subunit (left). Asn23 and Ser45 were changed to alanines, removing two hydrogen bonds (dashed lines) to biotin, and Ser27 was changed to aspartate, to introduce a steric clash. In monovalent streptavidin, the residues mutated in the dead subunits are shown in green (right). (c) To make monovalent streptavidin, dead streptavidin subunits (D) and wild-type streptavidin subunits (A) in a 3:1 ratio were refolded from denaturant, giving a mix of streptavidin heterotetramers. Tetramers with a single 6His-tagged wild-type subunit were purified on a Ni-NTA column. (d) SDS-PAGE of chimeric streptavidins under non-denaturing conditions. Streptavidin with 4 dead subunits (D4), wild- type streptavidin with a 6His-tag (A4), the product of refolding of D and A in a 3:1 ratio (Mix), and chimeric tetramers with one (A1D3), two (A2D2) or three (A3D1) biotin binding subunits were loaded without boiling onto a polyacrylamide gel and visualized by Coomassie staining. (e) SDS-PAGE of chimeric streptavidins under denaturing conditions to break the tetramer into monomers. Reproduced with permission from (Howarth et al., 2006).   184 5.6 CYTOSKELETON DEPOLYMERIZATION  Neurons were treated with 3mM nocodozole (NOCOD) or 4µM cytochalasin B (CytoB), pharmacological agents known to disrupt the microtubule and actin cytoskeleton respectively, through filament de-polymerization (Sabo and McAllister, 2003), prior to performing timelapse or FRAP experiments. This treatment did not affect the overall morphology of the neurons within the time period of the FRAP experiments (Appendix Figure 5.10).   Figure 5.10: Effect of CytoB and NOCOD treatment on the overall morphology of neurons. (A) AP-NL1 transfected neuron before CytoB treatment (left panel), and at the completion of ta FRAP experiment, 30 minutes after treatment (right panel). (B) AP-NL1 transfected neuron before NOCOD treatment (left panel), and at the completion of ta FRAP experiment, 30 minutes after treatment (right panel).               185 5.7 REFERENCES  Bresler, T., Ramati, Y., Zamorano, P. L., Zhai, R., Garner, C. C., and Ziv, N. E. (2001). The dynamics of SAP90/PSD-95 recruitment to new synaptic junctions. Mol Cell Neurosci 18, 149-167.  Bresler, T., Shapira, M., Boeckers, T., Dresbach, T., Futter, M., Garner, C. C., Rosenblum, K., Gundelfinger, E. D., and Ziv, N. E. (2004). Postsynaptic density assembly is fundamentally different from presynaptic active zone assembly. J Neurosci 24, 1507-1520.  El-Husseini, A. E., Schnell, E., Chetkovich, D. M., Nicoll, R. A., and Bredt, D. S. (2000). PSD-95 involvement in maturation of excitatory synapses. Science 290, 1364-1368.  Friedman, H. V., Bresler, T., Garner, C. C., and Ziv, N. E. (2000). Assembly of new individual excitatory synapses: time course and temporal order of synaptic molecule recruitment. Neuron 27, 57-69.  Howarth, M., Chinnapen, D. J., Gerrow, K., Dorrestein, P. C., Grandy, M. R., Kelleher, N. L., El-Husseini, A., and Ting, A. Y. (2006). A monovalent streptavidin with a single femtomolar biotin binding site. Nat Methods 3, 267-273.  Krueger, S. R., Kolar, A., and Fitzsimonds, R. M. (2003). The presynaptic release apparatus is functional in the absence of dendritic contact and highly mobile within isolated axons. Neuron 40, 945-957.  Okabe, S., Miwa, A., and Okado, H. (2001). Spine formation and correlated assembly of presynaptic and postsynaptic molecules. J Neurosci 21, 6105-6114.  Sabo, S. L., and McAllister, A. K. (2003). Mobility and cycling of synaptic protein-containing vesicles in axonal growth cone filopodia. Nat Neurosci 6, 1264-1269.  Shapira, M., Zhai, R. G., Dresbach, T., Bresler, T., Torres, V. I., Gundelfinger, E. D., Ziv, N. E., and Garner, C. C. (2003). Unitary assembly of presynaptic active zones from Piccolo-Bassoon transport vesicles. Neuron 38, 237-252.  Washbourne, P., Bennett, J. E., and McAllister, A. K. (2002). Rapid recruitment of NMDA receptor transport packets to nascent synapses. Nat Neurosci 5, 751-759.   186 6 Other Contributions  6.1 D1 RECEPTOR ACTIVATION INCREASES AMPA RECEPTOR SURFACE EXPRESSION AND SYNAPTIC RETENTION  A version of this section is under preparation. Gorlova N, Makado, Gerrow,K,  El-Husseini A, Philips A. (2008) D1 receptor activation increases AMPA receptor surface expression and synaptic retention independent of NMDA activity.  For this collaboration I performed the following experiments and analysis looking at the insertion of AP-tagged GluR1 upon treatment with the D1 agonist SKF81297. Activation of PKA by forscolin increases phosphorylation of Glur1 subunit of AMPA receptor at S845, and correlates with an increase surface expression of Glur1 subunit in hippocampal culture (Swayze et al., 2004).  In cortical and hippocampal cultures activation of DA D1 receptors leads to increase in surface expression of Glur1 subunit of AMPA receptor through the activation of  PKA (Gao et al., 2006).  However, direct evidence that phosphorylation of Glur1 at S845 is necessary for the D1-receptor evoked surface insertion of AMPA receptors is still lacking. In order to address this question, we used a surface biotinylation approach to monitor the insertion and surface localization of Glur1. Cortical cells were transfected with AP-tagged wild-type GluR1 subunit (AP-Glur1) or a phosphorylation-deficient GluR1 (AP-Glur1S845A), which can be recognized biotinylated by a bacterially derived biotin ligase, and visualized by incubation with fluorescently-conjugated streptavidin.  Intensity of GluR1 clusters was compared at sites apposed by synaptophysin (synaptic) and ³standing alone´ (non-synaptic).    Basal expression of AP-GluR1 was evident after 3-4 days of expression.  Cells exposed to 9 ȝM of SKF 81297 for 10 minutes showed significantly higher intensity of surface labeling AP-Glur1 at synaptic sites (4.96 ± 0.43 versus 2.97 ± 0.36, p=1.12e-9)  while there was no significant changes at extra-synaptic sites (1.95 ± 0.15 versus 1.86 ± 0.19 , p = 0.7, Appendix C Figure 1). In cells expressing AP-Glur1 S845A basal level of expression was significantly lower for both synaptic and extra-synaptic sites compared to cells expressing wild-type AP-GluR1 (1.03 ±0.13 versus  187 1.86±0.19, p=0.001; 1.76 ±0.20 versus 2.97 0.36, p=0.008 respectively). Compared to control, AP-Glur1 S845A expressing cells treated with SKF 81297 showed a significantly increase in intensity at extra-synaptic (1.65 ± 0.17 versus 1.03 ± 0.13, p = 0.005) and synaptic sites (2.54 ± 0.24 versus 1.75 ± 0.20, p = 0.01).  The degree of increase was significantly lower than in cells expressing wild type of AP-Glur1 (p=1.72e-05 at synaptic sites). Thus, stimulation of D1 receptors results in an increase of surface GluR1 containing AMPA receptors at the synapse. Mutation at S845 of Glur1 subunit significantly reduced but did not eliminated surface expression of this subunit, as well as, the enhanced surface expression by activation of D1 receptors.  Importantly, this mutation appears to affect the synaptic retention of GluR1 containing AMPA receptors.  Our findings are consistent with previous studies which demonstrated that activation of D1 receptors enhances AMPA GluR1 phosphorylation and surface expression through a PKA-mediated pathway (Wolf et al., 2003). Mutations of the PKA phosphorylation site of GluR1 subunit at Ser845 do prevent the delivery of the GluR1 subunit to synapses by active CaMKII or LTP (Lee et al., 2004; Roche et al., 1996). However, exogenous phosphorylation at this site does not induce the delivery of recombinant GluR1 subunit (Shi et al., 2001), indicating that PKA phosphorylation of the GluR1 subunit is necessary, but not sufficient, for its synaptic delivery. Potential linkers between PKA and the GluR1 subunit include the PKA-scaffolding molecule, AKAP, which binds to SAP97 and subsequently takes PKA to the vicinity of the GluR1 subunit (Colledge et al., 2000). The phosphorylation of Ser845 is correlated with a selective delivery of AMPARs to extrasynaptic sites (Ehlers, 2000; Oh et al., 2006)and the subsequent synaptic localization requires synaptic activation. Taken together, these data suggest a model in which Ser845 phosphorylation primes AMPARs for synaptic delivery by trafficking AMPARs first to the extra-synaptic membrane and then to the synapse via surface diffusion. The data presented herein, is consistent with this model, however, time-lapse experiments may provide the final key to describing this phenomena in full.  188   Figure 6.1: S845 in AP-GluR1 is important for the insertion and incorporation of AMPA receptors selectively at synaptic sites upon activation of D1 receptors. (A) Surface labeling of AP-GluR1 (green) in DIV 13 cortical neurons treated with control solution (left panels) or with SKF for 10 minutes (right panels). Cells were stained for synaptophysin (red) to denote synaptic regions. Magnified views of a representative dendrite (boxed regions) are in the lower panels. (B) Surface labeling of AP-GluR1 S845A (green) in DIV 13 cortical neurons treated with control solution (left panels) and with SKF for 10 minutes (right panels). Cells were stained for synaptophysin (red) to denote synaptic regions. Magnified views of a representative dendrite (boxed regions) are in the lower panels. (C) Quantification of GluR1 intensity at non-synaptic and synaptic sites. AP-GluR1 intensity at non-synaptic sites in control and SKF treated neurons did not show a significant enhancement (1.86 ± 0.19 versus 1.95 ± 0.15, p = 0.7), whereas there was a profound enhancement at synaptic sites (2.97 ± 0.36 versus 4.96 ± 0.43, p=1.12e-9). Under control conditions, AP-GluR1 S845A intensity at non- synaptic and synaptic sites was significantly decreased compared to wild type (1.86 ± 0.19 versus 1.03 ± 0.13, p = 0.001; 2.97 ± 0.36 versus 1.76 ± 0.20, p = 0.008 respectively). SKF treatment enhanced surface expression of AP-GluR1 S845A at both non-synaptic (1.03 ± 0.13 versus 1.65 ± 0.17, p = 0.005) and synaptic sites (1.75 ± 0.20 versus 2.54 ± 0.24, p = 0.01), albeit still significantly lower than wildtype AP-GluR1 (p=1.72e-05 at synaptic sites). Scale bar = 5 µm. *p<0.5. **p<0.01. ***p<.001  189 6.2 Als2-DEFICIENT MICE EXHIBIT DISTURBANCES IN ENDOSOMAL TRAFFICKING ASSOCIATED WITH MOTOR BEHAVIOURAL ABNORMALITIES.  A  version of this work has been published as: Devon RS, Orban PC, Gerrow K, Barbieri MA, Schwab C, Cao LP, Helm JR, Bissada N, Cruz-Aguado R, Davidson TL, Witmer J, Metzler M, Lam CK, Tetzlaff W, Simpson EM, McCaffery JM, El-Husseini AE, Leavitt BR, Hayden MR. (2006) Als2-deficient mice exhibit disturbances in endosome trafficking associated with motor behavioral abnormalities. Proc Natl Acad Sci U S A. Jun 20;103(25):9595- 600.  For this collaboration, I performed experiments and analysis the following two figures from the manuscript. We postulated that an impediment in endosomal trafficking of neurotrophin receptors, and resultant diminution in neurotrophic support, might be responsible for the reduction in neuronal size in Als2±/± mice. We analyzed cortical neurons and cerebellar granule cell neurons (CGNs) in culture, because both these cell types express alsin at high levels and cortical neurons in particular are affected in ALS. Neuronal morphology and survival in WT and Als2±/± neuron cultures were examined by DiI staining, fluorescence, and differential interference contrast (DIC) microscopy. No differences in morphology, size, or survival were noted between WT and Als2±/± neurons, and survival was between 85% and 95% for both genotypes (data not shown). We initially examined whether a reduction in Rab5-GEF activity caused by the absence of alsin may result in abnormal trafficking of neurotrophic factor receptors. After 3 h of trophic factor deprivation, cortical or CGNs at 8±12 days in culture were treated by bath application of 50 ng/ml BDNF for between 5 and 180 min, and the positions of the BDNF receptor, TrkB, were assessed by indirect immunofluorescence. In both cell types, there was a marked difference between WT and Als2±/± neurons in the accumulation of fluorescence in the cell bodies (Figure 6.2A, from cortical neurons). WT but not Als2±/± neurons showed a significant increase in perinuclear anti-TrkB staining in the first 60 min of BDNF stimulation (Figure 6.2B). There was no general disturbance of endocytosis, because uptake of Alexa Fluor 549-conjugated transferrin (Figure 6.2 C,D) or FITC- conjugated dextran (not shown) showed no difference between WT and Als2±/± cultures.  190    Figure 6.2: Neurons from Als2±/± mice show disturbances of BDNF receptor (TrkB) endocytosis. (A) Als2±/± cortical neurons show markedly less increase in fluorescence in the cell-body region after stimulation with BDNF than WT neurons. (B) Cell-body region TrkB fluorescence after BDNF stimulation, quantified in cortical neurons randomly selected by using differential interference contrast (DIC) microscopy (**, P < 0.01; 14±20 cells for each time point and genotype). (C) There is no difference in transferrin uptake between WT and Als2±/± neurons: after 30 min, both show intense transferrin fluorescence in their cell bodies. (D) Cell-body region fluorescence in a representative transferrin uptake time course, quantified in neurons randomly selected by using DIC microscopy. (Scale bars: 10 µm.)   We next repeated the BDNF stimulation experiment using insulin-like growth factor 1 (IGF1) and antibody to its receptor, IGF1R. After 30 min or more of IGF1 stimulation, puncta of at least 3.5 µm2 area (calculated after flattening the confocal images) were apparent in the dendrites of some (between 1/200 and 1/2,000) Als2±/± CGNs (Figure 6.3A). All positive cells contained at least 20 large puncta throughout several dendrites. Als2±/± cells were, on average, 1,000 times more likely to display this phenotype than WT cells (four separate cultures examined at the 60-min stimulation time point). To define the  191 nature of the IGF1R puncta, the cells were costained with markers of endocytic compartments. As shown in Figure 6.3, C and D, 32% of the puncta co-localized with early endosome antigen 1 (EEA1), and the majority (86%) co-localized with Rab5, both markers of early endosomes, and 17% co-localized with Rab11, a marker of recycling endosomes (Figure 6.3E). Similar results were obtained with cortical neurons, but a smaller proportion of the Als2±/± cells showed the phenotype (Devon et al., 2006).    Figure 6.3: Neurons from Als2±/± mice show disturbances of IGF1R endocytosis. (A) CGNs were stimulated with IGF1 for 0, 30, 60, and 360 min before staining with antibodies to the IGF1R alpha chain. Note the large puncta seen at later time points in Als2±/± neurons. (B) A cell with puncta is shown at lower magnification. (C±E) Als2±/± neurons, as in A, but stained with both anti-IGF1R and antibodies to EEA1 (C), Rab5 (D), or Rab11 (E). A large majority of IGF1R-positive puncta costain with anti-Rab5, a smaller proportion with anti-EEA1, and few with anti-Rab11. (Scale bars: 10 µm.)    192 6.3 A BALANCE BETWEEN EXCITATORY AND INHIBITORY SYNAPSES IS CONTROLLED BY PSD-95 AND NEUROLIGIN.   A version of following work has been published. Prange O, Wong TP*, Gerrow K*, Wang YT, El-Husseini A. (2004) A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. Proc Natl Acad Sci U S A. Sep 21;101(38):13915-20. *co-second authors   For this work I developed and performed over-expression experiments using contructs developed by O.Prange, who performed the analysis for these experiments. I developed, performed, and analyzed siRNA experiments looking at the E/I balance of synaptic inputs in neurons in which PSD-95 was reduced. Co-transfection of PSD-95 siRNA together with GFP reduced PSD-95 clusters (by 56 ± 5%). In contrast, no change in the number of PSD-95 clusters was observed in neurons expressing a scrambled siRNA (Figure 6.4 B). Expression of PSD-95 siRNA also resulted in an increase (1.5 ± 0.1-fold) in inhibitory (VGAT-positive) and a decrease (by 28 ± 7%) in excitatory (VGAT-negative) synaptic contacts. However, no change in total number of synaptic contacts was observed (Figure 6.4 C). These results demonstrate that the amounts of PSD-95 available can dictate the balance between excitatory and inhibitory presynaptic inputs (Prange et al., 2004).   193   Figure 6.4: Altered expression of PSD-95 influences the ratio of excitatory-to-inhibitory presynaptic contacts. (A) Hippocampal cells were transfected with either GFP (controls, Left) or PSD-95 GFP (Center) and immunostained at DIV 12 for GFP, synaptophysin (Syn), and the GABA inhibitory presynaptic marker VGAT. PSD-95 GFP expression resulted in a significant decrease in the percentage of VGAT-positive inhibitory contacts (Upper Right) but did not alter the number of PSD-95 clusters (Lower Right) (n = 15 cells, 288 VGAT+, 778 VGAT±) compared with controls (n = 12 cells, 242 VGAT+, 272 VGAT±). (B and C) Introduction of GFP with either a scrambled siRNA (control siRNA; Left) or PSD-95 siRNA (Center). (B Right) A reduction in the number of PSD-95-positive puncta in cells cotransfected with GFP and PSD- 95 siRNA (n = 11 cells, 301 puncta) compared with controls (n = 7 cells, 406 puncta) was found. (C Right) A significant increase in the percentage of inhibitory contacts (VGAT+) and a concurrent decrease in the percentage of excitatory contacts (VGAT±)(Upper) was found, but there was no change in the total number of synapses in cells expressing GFP and PSD-95 siRNA (n = 11 cells, 322 VGAT+, 282 VGAT±) compared with controls (n = 12 cells, 230 VGAT+, 494 VGAT±)(Lower). **, P < 0.01; ***, P < 0.001 (Mann±Whitney U test). (Scale bars: = 1 µm.)           194   6.4 NEUROLIGINS MEDIATE EXCITATROY AND INHIBITORY SYNAPSE FORMATION  A version of the following work has been published. Levinson JN, Chery N, Huang K, Wong TP, Gerrow K, Kang R, Prange O, Wang YT, El-Husseini A.  (2005) Neuroligins mediate excitatory and inhibitory synapse formation: involvement of PSD-95 and neurexin-1beta in neuroligin-induced synaptic specificity. J Biol Chem. Apr 29;280(17):17312-9.   For this work I performed and analyzed experiments looking at the distribution of NL1 and NL2 in young (DIV7) versus older hippocampal  neurons (DIV14) (Levinson et al., 2005).    6.5 PRESYNAPTIC TRAFFICKING OF SYNAPTOTAGMIN I  A version of the following work has been published. Kang R, Swayze R, Lise MF, Gerrow K, Mullard A, Honer WG, El-Husseini A. (2004) Presynaptic trafficking of synaptotagmin I is regulated by protein palmitoylation. J Biol Chem. Nov 26;279(48):50524-36.  For this work I performed the confocal microscopy and analysis to assess the surface retention of palmitoylation mutants of synaptotagmin I (Kang et al., 2004).  195 6.6  REFERENCES  Colledge, M., Dean, R. A., Scott, G. K., Langeberg, L. K., Huganir, R. L., and Scott, J. D. (2000). Targeting of PKA to Glutamate Receptors through a MAGUK-AKAP Complex. Neuron 27, 107-119.  Devon, R. S., Orban, P. C., Gerrow, K., Barbieri, M. A., Schwab, C., Cao, L. P., Helm, J. R., Bissada, N., Cruz-Aguado, R., Davidson, T.-L., et al. (2006). Als2-deficient mice exhibit disturbances in endosome trafficking associated with motor behavioral abnormalities. Proceedings of the National Academy of Sciences 103, 9595-9600.  Ehlers, M. D. (2000). Reinsertion or Degradation of AMPA Receptors Determined by Activity-Dependent Endocytic Sorting. Neuron 28, 511-525.  Gao, C., Sun, X., and W olf, M. E. (2006). Activation of D1 dopamine receptors increases surface expression of AMPA receptors and facilitates their synaptic incorporation in cultured hippocampal neurons. J Neurochem 98, 1664-1677.  Kang, R., Swayze, R., Lise, M. F., Gerrow, K., M ullard, A., Honer, W. G., and El-Husseini, A. (2004). Presynaptic trafficking of synaptotagmin I is regulated by protein palmitoylation. J Biol Chem 279, 50524- 50536.  Lee, S. H., Simonetta, A., and Sheng, M. (2004). Subunit Rules Governing the Sorting of Internalized AMPA Receptors in Hippocampal Neurons. Neuron 43, 221-236.  Levinson, J. N., Chery, N., Huang, K., Wong, T. P., Gerrow, K., Kang, R., Prange, O., Wang, Y. T., and El-Husseini, A. (2005). Neuroligins Mediate Excitatory and Inhibitory Synapse Formation: INVOLVEMENT OF PSD-95 AND NEUREXIN-1{beta} IN NEUROLIGIN-INDUCED SYNAPTIC SPECIFICITY. J Biol Chem 280, 17312-17319.  Oh, M. C., Derkach, V. A., Guire, E. S., and Soderling, T. R. (2006). Extrasynaptic Membrane Trafficking Regulated by GluR1 Serine 845 Phosphorylation Primes AMPA Receptors for Long-term Potentiation. J Biol Chem 281, 752-758.  Prange, O., Wong, T. P., Gerrow, K., Wang, Y. T., and El-Husseini, A. (2004). A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. PNAS 101, 13915-13920.  Roche, K. W., O'Brien, R. J., Mammen, A. L., Bernhardt, J., and Huganir, R. L. (1996). Characterization of Multiple Phosphorylation Sites on the AMPA Receptor GluR1 Subunit. Neuron 16, 1179-1188.  Shi, S.-H., Hayashi, Y., Esteban, J. A., and Malinow, R. (2001). Subunit-Specific Rules Governing AMPA Receptor Trafficking to Synapses in Hippocampal Pyramidal Neurons. Cell 105, 331-343.  Swayze, R. D., Lise, M.-F., Levinson, J. N., Phillips, A., and El-Husseini, A. (2004). Modulation of dopamine mediated phosphorylation of AMPA receptors by PSD-95 and AKAP79/150. Neuropharmacology 47, 764-778.  Wolf, M. E., Mangiavacchi, S., and Sun, X. (2003). Mechanisms by which dopamine receptors may influence synaptic plasticity. Ann N Y Acad Sci 1003, 241-249.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0066686/manifest

Comment

Related Items