Cellular and Molecular Mechanisms of Synaptic Specificity

University of Calgary PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2016 Cellular and Molecular Mechanisms of Synaptic Specificity

Getz, Angela

Getz, A. (2016). Cellular and Molecular Mechanisms of Synaptic Specificity (Unpublished doctoral thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/25237 http://hdl.handle.net/11023/3466 doctoral thesis

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY

Cellular and Molecular Mechanisms of Synaptic Specificity

Angela Michelle Getz

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

GRADUATE PROGRAM IN NEUROSCIENCE

CALGARY, ALBERTA

NOVEMBER, 2016

All nervous system functions rely upon the specificity of synapse formation and maturation, and

the experience-dependent remodeling of established synapses. Activity- and neurotrophic factor

(NTF)-dependent signaling events guide the appropriate expression and localization of various components of pre- and postsynaptic machinery to ensure synaptic network function is matched to the behavioral requirements of an animal. These processes govern the function of the individual synapses formed by any given neuron, although the underlying cellular and molecular mechanisms

have not been fully defined.

To determine postsynaptic mechanisms of synaptic specificity, I investigated the role of menin,

the product of the MEN1 tumor suppressor gene, in promoting cholinergic postsynaptic function

in response to NTFs. Here, I present the first direct evidence for the molecular actions of menin in

neurons, which coordinates the selective transcriptional upregulation and postsynaptic clustering

of neuronal nicotinic acetylcholine receptors (nAChR). This occurs through distinct actions of proteolytic fragments generated by activity-dependent calpain cleavage. These data identify a novel synaptogenic mechanism for the coordination of nuclear transcription and postsynaptic localization of neurotransmitter receptors by a single gene product, and identify menin as a candidate molecular scaffold for neuronal nAChR clustering.

To determine presynaptic mechanisms of synaptic specificity, I investigated how a co-transmitting neuron selectively employs classical or peptide neurotransmitters at synapses with distinct

postsynaptic targets. Here, I present the first evidence that the function of individual presynaptic

ii terminals is differentiated in a target cell-specific manner by an interplay between NTF and

retrograde arachidonic acid signaling. I found that presynaptic transmitter specificity was defined

by regulated synaptophysin expression, which selectively inhibited neuropeptide release

machinery. These observations identify a novel role for synaptophysin in the regulation of

peptidergic synaptic transmission, and a new component of NTF-dependent synaptic plasticity

through which the co-transmitter characteristics of individual presynaptic terminals are regulated.

Together, these studies delineate novel mechanisms underlying activity- and NTF-dependent

synaptic specificity, underscoring the importance of appropriate expression and localization of

synaptic machinery in controlling neuronal network function via the specialization of individual

synapses. These findings provide fundamental mechanistic insights into the neurodevelopmental,

neuropsychiatric, and neurodegenerative disorders in which these processes are disrupted.

iii Acknowledgements

To my supervisor, Dr. Naweed Syed – thank you for teaching me to be an experimentalist, to find wonder and humility in science, and for always giving me the freedom and encouragement to explore whatever observation peaked my curiosity.

To my committee members, Dr. Gerald Zamponi and Dr. Michael Colicos – thank you for providing the guidance, experimental insights, and fresh perspectives necessary to develop and realize these ideas.

To Wali – thank you for your tireless dedication to the craft, and for always making my increasingly demanding cell culture requests a reality.

To Frank – thank you for teaching me everything I know about molecular biology, and for always encouraging me to take a step back, and then try something new.

To the past and present members of the Syed lab, especially Fenglian, Collin, Tara, Pierre, Svetlana and Jean – thank you for your collaboration and friendship along the way, for sharing in the excitement and frustration that comes with science, and for always making day to day life in the lab amusing.

iv To my parents, Susan and Darrell – thank you for your endless support, for encouraging me to pursue my dreams even when it means moving away from you, and for both uniquely prompting a lifelong curiosity in how and why the brain does what it does, which has given me direction in life.

To my incredible family and friends back home in Regina – thank you for giving me perspective in life, and for always being there for me, no matter the distance.

To the wonderful friends I have made during my time in Calgary – thank you for all the good times over the past decade, and for making this city my new home.

To Tobi – thank you for being a true partner in life, and for your limitless curiosity and sense of adventure that parallels my own. Thank you for your patience and understanding, for your support and encouragement, and for always walking the dog when I was feeling overwhelmed while writing up.

v Table of Contents

Abstract ...... ii Acknowledgements ...... iv Table of Contents ...... vi List of Tables ...... x List of Figures and Illustrations ...... xii List of Symbols, Abbreviations and Nomenclature ...... xv Epigraph ...... xxi

CHAPTER ONE: INTRODUCTION ...... 1

1.1 General Introduction ...... 1 1.2 Synaptic Transmission ...... 3 1.2.1 Chemical synapses: neurotransmitters and neurotransmitter receptors ...... 3 1.2.2 Chemical synapses: specificity of transmission at the presynaptic terminal .....5 1.2.3 Chemical synapses: specificity of transmission at the postsynaptic terminal ...8 1.3 Synapse Formation ...... 10 1.3.1 Intrinsic signaling: the ‘ready-set-go’ model of synaptogenesis ...... 10 1.3.2 Cell-cell signaling: cell adhesion molecules induce synaptic differentiation .11 1.3.3 Cell-cell signaling: neurotransmitter-receptor interactions ...... 13 1.3.3.1 Cell-cell signaling: ‘black-box’ factors – peptide neurotransmitter- receptor interactions ...... 15 1.3.4 Extrinsic signaling: neurotrophic factors ...... 16 1.3.5 Extrinsic regulation of intrinsic signaling: genetic foundations of synaptogenesis ...... 18 1.3.5.1 Activity-dependent gene expression ...... 19 1.3.5.2 NTF-dependent gene expression ...... 21 1.3.5.3 Signal integration: coincidence detection ...... 22 1.3.5.4 Signal integration: ‘black-box’ factors – the MEN1 gene ...... 23 1.4 Synaptic Specificity ...... 25 1.4.1 Cell-cell signaling: the ‘chemoaffinity’ model of synaptic specificity ...... 25 1.4.2 Signal integration: mechanisms that specify the function of individual synapses ...... 26 1.4.2.1 Cell-cell signaling: NTF-dependent interactions ...... 26 1.4.2.2 Cell-cell signaling: activity- and neurotransmitter-receptor dependent interactions ...... 27 1.4.2.3 Cell-cell signaling: ‘black-box’ factors – presynaptic transmitter specificity ...... 29 1.5 Specific background and rationale of the Lymnaea model system ...... 33 1.5.1 Synaptic cell culture: the soma-soma model ...... 34 1.5.2 The Lymnaea model: nicotinic acetylcholine receptors ...... 35 1.5.3 The Lymnaea model: FMRFamide and peptidergic synaptic transmission ....36 1.6 Specific Aims ...... 38

vi CHAPTER TWO: TWO PROTEOLYTIC FRAGMENTS OF MENIN COORDINATE THE NUCLEAR TRANSCRIPTION AND POSTSYNAPTIC CLUSTERING OF NEUROTRANSMITTER RECEPTORS DURING SYNAPTOGENESIS BETWEEN LYMNAEA NEURONS ...... 44

2.1 Abstract ...... 44 2.2 Introduction ...... 46 2.3 Methods ...... 48 2.3.1 Animals and neuronal cell culture ...... 48 2.3.2 Molecular biology ...... 48 2.3.3 Immunocytochemistry and microscopy ...... 49 2.3.4 Preparation of protein samples and Western blotting ...... 50 2.3.5 Electrophysiology ...... 52 2.3.6 Chemicals ...... 52 2.3.7 Experimental design and statistical analysis ...... 53 2.4 Results ...... 54 2.4.1 Menin is localized to a synapse between Lymnaea neurons ...... 54 2.4.2 Menin is cleaved by calpain and the resulting fragments are differentially localized within neurons ...... 58 2.4.3 Postsynaptic recruitment of the C-menin fragment requires neurotrophic factor- and activity-dependent signaling ...... 69 2.4.4 The C-menin fragment mediates postsynaptic consolidation of excitatory nAChR ...... 73 2.4.5 Menin fragments coordinate subunit-specific transcriptional upregulation and synaptic targeting of excitatory nAChR ...... 81 2.5 Discussion ...... 87 2.5.1 Proteolytic cleavage ...... 87 2.5.2 Transcriptional regulation ...... 88 2.5.3 Postsynaptic recruitment ...... 90 2.5.4 Conclusion ...... 92

CHAPTER THREE: TUMOR SUPPRESSOR MENIN IS REQUIRED FOR SUBUNIT-SPECIFIC NACHR α5 TRANSCRIPTION AND NACHR- DEPENDENT PRESYNAPTIC FACILITATION IN CULTURED MOUSE HIPPOCAMPAL NEURONS ...... 102

3.1 Abstract ...... 102 3.2 Introduction ...... 104 3.3 Methods ...... 108 3.3.1 Animals and neuronal cell culture ...... 108 3.3.2 Molecular biology ...... 109 3.3.3 Lentivirus production and transduction of neuronal cultures ...... 109 3.3.4 Immunocytochemistry and microscopy ...... 110 3.3.5 Electrophysiology ...... 112 3.3.6 Experimental design and statistical analysis ...... 112

vii 3.4 Results ...... 114 3.4.1 Menin proteolytic cleavage fragments are differentially localized within neurons ...... 114 3.4.2 C-menin co-localizes with α7 subunit-containing nAChRs at glutamatergic presynaptic terminals ...... 125 3.4.3 Menin mediates subunit-selective transcriptional regulation of nAChR α5 .131 3.4.4 Menin is required for the functional expression of nAChRs in hippocampal neurons ...... 137 3.4.5 MEN1 transcription and menin proteolytic cleavage are disrupted in a mouse model of Alzheimer’s disease ...... 143 3.5 Discussion ...... 147 3.5.1 Menin proteolytic cleavage and differential subcellular distribution ...... 147 3.5.2 Menin-dependent transcriptional regulation ...... 148 3.5.3 nAChRs on presynaptic terminals: subunit composition, function, and targeting ...... 150 3.5.4 The role of menin and nAChRs in synapse formation, plasticity, and maintenance ...... 153

CHAPTER FOUR: A NOVEL MECHANISM OF SYNAPTIC SPECIFICITY: SYNAPTOPHYSIN INHIBITS PRESYNAPTIC SECRETORY MACHINERY TO REGULATE THE RELEASE OF PEPTIDE NEUROTRANSMITTERS...... 163

4.1 Abstract ...... 163 4.2 Introduction ...... 164 4.3 Methods ...... 167 4.3.1 Animals and neuronal cell culture ...... 167 4.3.2 Molecular biology ...... 167 4.3.3 Immunocytochemistry, immunohistochemistry, in situ hybridization and imaging ...... 168 4.3.4 Electrophysiology ...... 169 4.3.5 Chemicals ...... 169 4.3.6 Experimental design and statistical analysis ...... 170 4.4 Results ...... 171 4.4.1 Postsynaptic target cell specificity and extrinsic neurotrophic factors regulate peptidergic synaptic transmission ...... 171 4.4.2 Presynaptic inhibition of neuropeptide secretory machinery defines NTF- and target cell-dependent presynaptic transmitter specificity ...... 177 4.4.3 Presynaptic expression of synaptophysin is regulated by neurotrophic factors and modulated by postsynaptic identity ...... 180 4.4.4 Synaptophysin inhibits neuropeptide release ...... 192 4.4.5 Trans-synaptic retrograde signaling by arachidonic acid metabolites regulates synaptophysin expression and presynaptic peptide transmitter specificity ...... 195 4.4.6 Cell type-specific profiles of synaptophysin expression ...... 199

viii 4.4.7 Presynaptic peptide transmitter specificity is established notwithstanding competing molecular signals from distinct postsynaptic targets ...... 204 4.5 Discussion ...... 214 4.5.1 Differential sorting and trafficking of large dense-core vesicles ...... 215 4.5.2 Context-dependent posttranslational modifications of synaptophysin ...... 215 4.5.3 Extracellular and cell-cell signaling in peptidergic synaptic specificity ...... 217 4.5.4 Functional implications of synaptophysin-regulated peptidergic transmission ...... 218

CHAPTER FIVE: GENERAL DISCUSSION ...... 232

5.1 Summary of results ...... 232 5.2 Significance of findings, identification of limitations, and future directions ...... 236 5.2.1 The role of MEN1/menin in postsynaptic specificity ...... 236 5.2.2 The role of neuropeptides and synaptophysin in presynaptic specificity ...... 248 5.2.3 Implications for synaptic network function ...... 255 5.2.3.1 Modeling behaviour in Lymnaea by the synaptic properties of individual neurons ...... 255 5.2.3.2 Potential applications to human health and disease ...... 256 5.2.4 General conclusion ...... 260

CHAPTER SIX: APPENDIX I: OVERVIEW OF LYMNAEA EXPERIMENTAL PREPARATIONS ...... 261

6.1 Synaptic cell culture ...... 261 6.2 Single cell qPCR ...... 264

CHAPTER SEVEN: APPENDIX II: ACKNOWLEDGEMENTS AND CONTRIBUTIONS ...... 268

REFERENCES ...... 282

ix List of Tables

Table 1.1 - Cellular and molecular mechanisms underlying synaptic specificity ...... 42

Table 1.2 - A partial list of presynaptic proteins and their functions ...... 43

Table 2.1 - Lymnaea cloning primers ...... 93

Table 2.2 - Lymnaea RT-PCR gene specific primers ...... 94

Table 2.3 - Lymnaea qPCR gene specific primers ...... 95

Table 2.4 - Immunocytochemistry fluorescence of menin in LPeD1 ...... 96

Table 2.5 - Immunocytochemistry fluorescence of c-Myc-tagged menin in LPeD1 ...... 96

Table 2.6 - Subcellular distribution of menin in VD4-LPeD1 axon-axon pairs ...... 97

Table 2.7 - Incidence of VD4-LPeD1 excitatory synapse formation and EPSP amplitudes ...... 98

Table 2.8 - EPSP amplitudes of VD4-LPeD1 synapses and incidence of excitatory nAChR expression in single LPeD1 neurons ...... 99

Table 2.9 - Relative gene expression in LPeD1 neurons ...... 100

Table 3.1 - Mouse qPCR gene specific primers (intron spanning) ...... 156

Table 3.2 - Immunocytochemistry fluorescence of menin in hippocampal neuron cultures ...... 157

Table 3.3 - Incidence of C-menin co-localization in super-resolved immunocytochemistry images ...... 157

Table 3.4 - Relative gene expression in hippocampal neuron cultures ...... 158

Table 3.5 - Immunocytochemistry fluorescence of menin and nAChRs in NTC shRNA- transduced hippocampal neuron cultures ...... 159

Table 3.6 - Immunocytochemistry fluorescence of menin and nAChRs in MEN1 shRNA- transduced hippocampal neuron cultures ...... 160

Table 3.7 - mEPSC amplitude and frequency values in hippocampal neuron cultures ...... 161

Table 3.8 - Gene expression in 5xFAD +/- relative to 5xFAD -/- hippocampus ...... 162

Table 3.9 - Western blot analysis of menin in 5xFAD +/- and 5xFAD -/- hippocampus ...... 162

Table 4.1 - Lymnaea cloning primers ...... 220

Table 4.2 - Lymnaea RT-PCR gene specific primers ...... 220

x Table 4.3 - Lymnaea qPCR gene specific primers ...... 221

Table 4.4 - Lymnaea in situ hybridization probe sequences ...... 222

Table 4.5 - Peptidergic characteristics of VD4-VF and VD4-RPeD1 synapses ...... 222

Table 4.6 - Somatic FMRFa fluorescence in VD4 ...... 223

Table 4.7 - Synaptic FMRFa fluorescence ...... 223

Table 4.8 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in VD4 CM ...... 224

Table 4.9 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in VD4 DM ...... 225

Table 4.10 - Single cell qPCR - Normalized expression of G proteins in VF ...... 226

Table 4.11 - Single cell qPCR - Normalized expression of G proteins in RPeD1 ...... 226

Table 4.12 - ACh and FMRFa synaptic transmission characteristics of VD4-VF pairs ...... 227

Table 4.13 - Peptidergic characteristics of VD4-RPeD1 synapses ...... 228

Table 4.14 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in VD4 ... 229

Table 4.15 - Peptidergic characteristics of VD4-VF synapses in DM ...... 230

Table 4.16 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in peptidergic cell types ...... 230

Table 4.17 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in non- peptidergic cell types ...... 231

Table 4.18 - Synaptic FMRFa fluorescence in RPeD1-VD4-VF triple axon pairs in CM ...... 231

Table 4.19 - Peptidergic characteristics of RPeD1-VD4-VF synapses in vitro and in vivo ...... 231

Table 6.1 - Identified neurons employed in this thesis for the study of synaptic specificity ...... 261

xi List of Figures and Illustrations

Figure 1.1- Overview of specific aims and hypotheses ...... 40

Figure 2.1 - Postsynaptic clustering of menin in Lymnaea neurons ...... 56

Figure 2.2 - Menin is cleaved by calpain ...... 62

Figure 2.3 - The menin conserved region of interest ...... 64

Figure 2.4 - Predicted calpain cleavage site in the menin region of interest is conserved ...... 66

Figure 2.5 - Menin fragments are differentially targeted within LPeD1 neurons ...... 68

Figure 2.6 - Postsynaptic recruitment of C-menin requires neurotrophic factor- and activity- dependent signaling ...... 71

Figure 2.7 - The C-menin fragment mediates postsynaptic consolidation of excitatory nAChRs ...... 76

Figure 2.8 - Synthetic mRNA-induced menin expression is specific to microinjected neurons .. 78

Figure 2.9 - The C-menin fragment is required for postsynaptic consolidation but not functional expression of excitatory nAChRs ...... 79

Figure 2.10 - Menin fragments coordinate subunit-specific transcriptional upregulation and synaptic targeting of excitatory nAChRs ...... 84

Figure 2.11 - A model for the coordination of nuclear transcription and postsynaptic clustering of excitatory nAChRs by menin proteolytic fragments ...... 86

Figure 3.1 - N- and C-terminal epitopes recognized by the menin antibodies ...... 117

Figure 3.2 - Menin fragments are differentially localized in CNS neurons ...... 118

Figure 3.3 - Menin exhibits neuron-specific expression in hippocampal cultures ...... 120

Figure 3.4 - Menin is cleaved by calpain ...... 122

Figure 3.5 - Menin proteolytic fragments are present in non-neuronal tissue ...... 124

Figure 3.6 - The C-terminal menin fragment co-localizes with α7 subunit-containing nAChRs at presynaptic terminals ...... 127

Figure 3.7 - Field of view images for super-resolution microscopy ...... 129

Figure 3.8 - Menin mediates subunit-specific transcriptional regulation of nAChR α5 ...... 133

Figure 3.9 - Menin knockdown reduces nAChR α5 expression ...... 135

xii Figure 3.10 - MEN1 knockdown eliminates nAChR-dependent presynaptic facilitation ...... 139

Figure 3.11 - Control traces for patch clamp recordings ...... 141

Figure 3.12 - Histogram distribution of mEPSC amplitudes in mouse hippocampal neurons ... 142

Figure 3.13 - MEN1 transcription and menin proteolytic cleavage are disrupted in the 5xFAD mouse model of Alzheimer's disease ...... 145

Figure 4.1 – Presynaptic transmitter specificity is regulated by neurotrophic factors and postsynaptic target identity ...... 173

Figure 4.2 - Postsynaptic receptor expression profiles ...... 175

Figure 4.3 - Inhibition of neuropeptide release defines presynaptic transmitter specificity ...... 178

Figure 4.4 - Prediction of Lymnaea synaptophysin transmembrane domains ...... 183

Figure 4.5 - Sequences of Lymnaea synaptophysin splice variants ...... 185

Figure 4.6 - Sequences of Lymnaea synaptotagmin I C2B-α and C2B-β splice variants ...... 187

Figure 4.7 - Prediction of Lymnaea synaptotagmin I C2B-α transmembrane domain ...... 188

Figure 4.8 - Presynaptic synaptophysin expression is regulated by neurotrophic factors and postsynaptic target cell identity...... 190

Figure 4.9 - Synaptophysin inhibits neuropeptide release ...... 193

Figure 4.10 - Peptide transmitter specificity is induced by arachidonic acid metabolite retrograde signaling ...... 197

Figure 4.11 - Cell type-specific profiles of synaptophysin expression ...... 201

Figure 4.12 - In situ hybridization in the Lymnaea CNS ...... 203

Figure 4.13 - Peptide transmitter specificity occurs with competing postsynaptic targets ...... 206

Figure 4.14 - Peptide transmitter specificity occurs in vivo despite elevated synaptophysin expression ...... 208

Figure 4.15 - A model for neurotrophic factor and cell-cell signaling in synaptophysin- dependent presynaptic peptide transmitter specificity ...... 210

Figure 4.16 - A model for posttranslational modification of synaptophysin in synapse- specific inhibition of neuropeptide release ...... 212

Figure 5.1 – A potential mechanism for neuronal nAChR clustering by coiled-coil domains ... 247

xiii Figure 5.2 - Proposed model for the clinical relevance of synaptic specificity mechanisms ..... 259

Figure 6.1 - Lymnaea cell culture preparations ...... 262

Figure 6.2 - Lymnaea single cell qPCR ...... 266

xiv List of Symbols, Abbreviations and Nomenclature

5-HT Serotonin 5xFAD Transgenic mouse line expressing 5 familial AD gene mutations α-BTX α-Bungarotoxin; nAChR α7 inhibitor AA Arachidonic acid metabolites AC Adenylyl Cyclase ACh Acetylcholine AD Alzheimer’s disease AMPA α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor ANOVA Analysis of variance APP Amyloid precursor protein

APV 2-amino-5-phosphonovalerate; NMDA receptor antagonist

ARIA Acetylcholine receptor inducing activity

ATP Adenosine triphosphate AU Arbitrary units; measure of fluorescence intensity AZ Active zone βACT β-actin βTUB β-tubulin BDNF Brain-derived neurotrophic factor Ca2+ Calcium ion CAM Cell adhesion molecule CaMKII Ca2+/calmodulin-dependent protein kinase II cAMP cyclic adenosine monophosphate CASK Ca2+/calmodulin-dependent serine protein kinase CAST Cytomatrix associated structural protein CAZ Cytoplasmic matrix of the active zone CDC Cerebral caudodorsal cell cDNA Complementary DNA Ch Cl Chelerythrine chloride; PKC inhibitor

xv CKO Conditional knockout CM CNS-conditioned media; NTF-rich C-MEN1 C-terminal MEN1 mRNA fragment C-menin C-terminal menin fragment CNS Central nervous system CNQX 6-cyano-7-nitroquinoxaline-2,3-dione; AMPA receptor antagonist CoDiM Conical diffraction microscopy (super resolution technique) CPLX Complexin CRE cAMP response element CREB CRE binding protein C-terminus carboxyl terminus (protein) DG Dentate gyrus DIV Days in vitro Dlg Discs large (gene) DM Defined media; NTF-deficient DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid EGF Epidermal growth factor eGFP Enhanced green fluorescent protein EGFR Epidermal growth factor receptor ELH Egg laying hormone EPSP Excitatory postsynaptic potential ER Endoplasmic reticulum EST Expressed sequence tag ETYA 5,8,11,14-Eicosatetraynoic acid; LOX inhibitor FaRP FMRFamide related peptide FGF Fibroblast growth factor

FMRFa FMRFamide (Phe-Met-Arg-Phe-NH2); also refers to the various neuropeptides derived from FMRFamide gene transcripts GABA γ-aminobutyric acid GFAP Glial fibrillary acidic protein

xvi GKAP Guanylate kinase-associated protein GPCR G protein coupled receptor GRIP Glutamate receptor-interacting protein GSK3β Glycogen synthase kinase 3 beta GSP Gene specific primer HA Hemagglutinin (epitope tag) HEK-293 Human embryonic kidney 293 cell line HH3 Histone H3 Hz Hertz (per second); frequency measurement unit ICC Immunocytochemistry IHC Immunohistochemistry IM Incubation media ISH In situ hybridization kDa KiloDalton Kv voltage-gated potassium channel LDCV Large dense-core vesicle L-MEN1 Lymnaea MEN1 L-menin Lymnaea menin L-NNA NG-Nitro-L-arginine; NOS inhibitor LOX Lipoxygenase LPeD1 Left pedal dorsal 1 LTD Long-term depression LTF Long-term facilitation LTP Long-term potentiation MAPK Mitogen-activated protein kinase MECP2 Methyl CpG binding protein 2 (gene) MEN1 Multiple endocrine neoplasia type 1 (gene) mEPSC Miniature excitatory postsynaptic current MLA Methyllcaconitine; AChR antagonist MLL1 Mixed-lineage leukemia 1 (gene) MN Motoneuron

xvii mo Months old mRNA Messenger RNA MuSK Muscle specific kinase mV Millivolt; membrane potential measurement unit Myc c-Myc (epitope tag) nAChR Nicotinic ACh receptor NBM Neurobasal media NES Nuclear exit signal NeuN Neuronal nuclei; Fox3 (gene) NFκβ Nuclear factor kappa-light-chain-enhancer of activated B cells NGF Nerve growth factor NLS Nuclear localization signal NMDA N-methyl-D-aspartate receptor N-MEN1 N-terminal MEN1 mRNA fragment N-menin N-terminal menin fragment NMJ Neuromuscular junction NO Nitric oxide NOS Nitric oxide synthase NPY Neuropeptide Y NSF N-ethylmaleimide sensitive fusion protein NT-3 Neurotrophin 3 N-terminus Amino terminus (protein) NTF Neurotrophic factor pA Picoamp; current measurement unit PDZ PSD-95/Dlg1/zo-1 domain PI3K Phosphatidylinositol 3 kinase PMA Phorbol 12-myristate 13-acetate; PKC activator PC12 Pheochromocytoma cell line PCR Polymerase chain reaction PeA Pedal A cluster neuron PICK1 Protein interacting with C kinase

xviii PKA Protein kinase A PKC Protein kinase C PLA2 Phospholipase A2 PLC Phospholipase C PSD Postsynaptic density PSD-95/93 Postsynaptic density protein 95/93 (referring to 95 or 93 kDa) PS1 Presenilin1 PSI Pounds per square inch; pressure measurement unit PSP Postsynaptic potential PTEN Phosphatase and tensin homolog (gene) PTM Posttranslational modification PTV Piccolo-Bassoon transport vesicle

PTX Pertussis toxin; Gαi/o inhibitor qPCR Quantitative PCR rCPG Respiratory central pattern generator (CPG) REST Relative expression software tool RIM Rab3 interacting molecule ROI Region of interest RNA Ribonucleic acid RNAi RNA interference RPeD1 Right pedal dorsal 1 rRNA ribosomal RNA RT Room temperature / Reverse transcription (PCR) RTK Receptor tyrosine kinase SEM Standard error of the mean shRNA small-hairpin RNA; a method of RNAi SN Sensory neuron SNAP Soluble NSF attachment protein SNARE SNAP receptor SSV Small synaptic vesicle STF Short-term facilitation

xix STXBP1 Syntaxin binding protein 1; Munc18 (gene) SV Synaptic vesicle SVP Synaptic vesicle protein Syb Synaptobrevin Syp Synaptophysin Syt Synaptotagmin Syt-α Synaptotagmin I C2B-α Syt-β Synaptotagmin I C2B-β Syx Syntaxin TC Tubocurarine; AChR antagonist TEA Tetraethylammonium; AChR antagonist Trk Tropomyosin receptor kinase; neurotrophin receptor t-SNARE Target-localized SNARE TTX Tetrodotoxin; voltage-gated sodium channel antagonist VD4 Visceral dorsal 4 interneuron VF Viseral F group neurons VGCC Voltage-gated Ca2+ channel v-SNARE vesicle-localized SNARE WB Western blot

xx Epigraph

If the human brain were so simple that we could understand it, we would be so simple

that we couldn’t

- E.M. Pugh, as quoted in “The Biological Origin of Human Values” by G.E. Pugh

xxi

Chapter One: Introduction

1.1 General Introduction

All functions of the nervous system are contingent upon the precise organization of neuronal connections that are initially patterned during development, and then continually modified throughout life. Communication between neurons is mediated by synapses, specialized cell-cell contact sites characterized by asymmetrical oppositions of electron-dense membrane thickenings, the ‘presynaptic active zone’ and the ‘postsynaptic density’, separated by a synaptic cleft. Synaptic vesicles (SVs) containing neurotransmitter substances are clustered nearby or functionally docked by interactions with scaffolds of the cytoplasmic matrix of the active zone (CAZ). SV fusion at the presynaptic active zone (AZ) is coupled to calcium (Ca2+) influx triggered by action potential depolarizations and the opening of voltage-gated Ca2+ channels (VGCC). Neurotransmitter substances released from SVs diffuse across the cleft and bind receptor macromolecules that are embedded within the membrane and linked to an extensive cytoplasmic scaffold that comprises the postsynaptic density (PSD). Receptor activation generally influences the membrane potential of the postsynaptic neuron in a positive (excitatory) or negative (inhibitory) direction, and thus information is transmitted from one neuron to the next1,2. The selective formation, refinement, plastic remodeling and retention of synapses between individual neurons is a remarkably specific phenomenon. A number of the cellular and molecular mechanisms that define these processes, however, remain undefined. Determining the mechanisms that produce synaptic specificity is critical to advancing both our fundamental understanding of the nervous system, as well as the various neurodevelopmental, neurological, neuropsychiatric, or neurodegenerative disorders that

are met in clinical practice when these processes go awry (e.g. autism, epilepsy, depression,

schizophrenia, Alzheimer’s disease).

Synaptic specificity is a broad and often vague area of neuroscience in which there are currently

more questions than answers: How is a neuron able to selectively generate the appropriate and

diverse types of presynaptic outputs or postsynaptic responses that underlie the functional

flexibility of neuronal networks? How does a neuron establish specific types of synapses with

distinct targets? How is a neuron able to differentiate the molecular composition and function of

its individual pre- and postsynaptic terminals? What are the mechanisms by which such specificity

is established during synapse formation or maintained at mature synapses? Can synaptic specificity

be altered by experience, and if so, what defines this form of plasticity? How do the properties of

synaptic specificity influence information processing within neuronal networks? What are the

functional consequences to behaviour that arise when synaptic specificity is perturbed? In order to

understand the incredible structural and functional complexity of the brain and its capacity for

flexibility, it is imperative to elucidate the cellular and molecular mechanisms responsible for

establishing the specificity of individual pre- and postsynaptic structures. This chapter will provide

a literary overview of what is currently known about the cellular and molecular mechanisms

underlying how synaptic specificity emerges (Table 1.1), via regulated synaptic transmission

(Section 1.2), the selective development and refinement of synaptic connections (Section 1.3), and the differentiation of individual synapses as a result of context-dependent signals that influence target recognition, synaptic maturation and plasticity (Section 1.4).

1.2 Synaptic Transmission

1.2.1 Chemical synapses: neurotransmitters and neurotransmitter receptors

A surprising number of neurotransmitter substances are employed by neurons, and co-

transmission, the use of multiple transmitters by a single neuron, is now generally considered to

be the rule rather than the exception. There are currently three recognized classes of

neurotransmitters: (i) classical small molecule transmitters that are synthesized enzymatically,

such as acetylcholine (ACh), glutamate, or γ-aminobutyric acid (GABA); (ii) peptide transmitters that are synthesized by the proteolytic processing of larger precursor proteins, such as the invertebrate neuropeptide FMRFamide or the mammalian neuropeptide Y (NPY); and (iii) atypical transmitters, such as ATP, arachidonic acid metabolites (AA), or nitric oxide (NO). Classical transmitters are stored in small synaptic vesicles (SSVs; ~40-50 nm diameter) that are synthesized, filled, and recycled at presynaptic terminals, whereas peptide neurotransmitters are synthesized within large dense-core vesicles (LDCVs; ~80-100 nm) that are typically derived from the endoplasmic reticulum (ER)/Golgi network, trafficked to presynaptic terminals, and do not recycle. The highly heterogeneous group of atypical transmitters represents a unique case in that they can be co-stored in SVs with other neurotransmitters (e.g. ATP), or synthesized ‘on-demand’ and released without vesicular packaging due to their highly membrane-permeable nature (e.g.

AA, NO)3. The use of atypical transmitters such as NO or AA also allows for retrograde synaptic

transmission, by which postsynaptic activity can transmit information presynaptically. The

molecular mediators and functional implications of this form of synaptic communication are not

yet fully understood, although recent investigations are identifying surprising targets, such as the

direct inhibition of presynaptic voltage-gated potassium channels (Kv) by AA at mossy fiber

synapses, which produces synaptic facilitation by the broadening of the presynaptic action potential4.

Even more complex than the diversity of neurotransmitters are the postsynaptic receptors that

receive these signals across the synapse. Trans-synaptic communication is primarily effected by

signal transduction through ionotropic (fast-acting; ligand-gated ion channels) or metabotropic receptors (slow-acting; heterotrimeric G protein coupled receptors [GPCR]). Most types of classical neurotransmitter substances signal through both ionotropic and metabotropic receptors, whereas most neuropeptides signal exclusively through metabotropic receptors. Ionotropic receptors are multimeric protein complexes typically comprised of heterologous subunits that exhibit variable expression levels in different cell types and/or across development, each of which confers a different functional property to the channel, such as ligand binding affinity, gating kinetics, ion conductance or permeability (especially variations in Ca2+), or trafficking. For

instance, mammalian neuronal nicotinic acetylcholine receptors (nAChR) are pentameric channels

comprised of different combinations of subunits (α2-10, β2-4) that exhibit distinct patterns of expression across development and in different brain regions5. The nAChRα5 subunit in particular

confers on the channel the properties of enhanced ACh sensitivity and Ca2+ conductance, and its

upregulation during the plastic remodeling of the spinal cord following peripheral nerve injury has

been found to potentiate cholinergic synaptic transmission and contribute to the development of

neuropathic pain6-8. Metabotropic receptors, on the other hand, are singular proteins with 7

characteristic transmembrane domains that are coupled to heterotrimeric G protein complexes,

linking extracellular signaling events to the generation of intracellular second messengers. Like

ionotropic receptors, metabotropic receptors exhibit distinct expression patterns and functional

variability through coupling with different types of G protein subunits. The invertebrate

neuropeptide FMRFamide, for example, is known to elicit either excitatory or inhibitory

postsynaptic responses in a cell-type specific manner. Although the FMRFamide GPCR is yet to

be identified, these findings are indicative of differential coupling of the receptor, presumably by

the expression of different isoforms, to distinct classes of G proteins9,10. The expression of

particular types of receptors and the generation of particular postsynaptic responses, whether it be

a cell type-specific phenomenon or induced by context-dependent plasticity such as injury

remodeling, illustrates one of the mechanisms by which synaptic specificity can emerge.

1.2.2 Chemical synapses: specificity of transmission at the presynaptic terminal

Neurotransmitter release by the exocytosis of SVs is the primary mechanism through which

information is relayed in neuronal networks. The molecular components of the CAZ and the SV

release machinery are summarized in Table 1.211. The characteristics of presynaptic transmitter

release are defined by several interdependent variables: the Ca2+ microdomain; the properties of

the SV protein (SVP) isoforms present and their Ca2+ binding capacity, phosphorylation or

glycosylation states; physical associations amongst SVPs; and the dynamics of the fusion pore1.

All of these factors are highly variable, and important sources of presynaptic functional diversity.

While the contributions of many SVPs to transmitter release are well defined (e.g. synaptotagmin

I, synaptobrevin 2, synapsin I)12-14, the roles of others are less clear because their influence on

synaptic transmission is subtle. The presence of functional redundancy due to extensive isoform

variations (e.g. 17+ isoforms of synaptotagmin in mammals) often further confounds the role of

each SVP, although work aimed at deciphering this complexity is in progress. For instance,

synaptotagmin 7 has recently been identified as a specialized calcium sensor for use-dependent facilitation at mossy fiber synapses, a form of short-term synaptic plasticity dependent upon residual Ca2+ in which subsequent action potentials enhance SV release15. A particularly intriguing

case of SVP functional specialization that remains obscure is the synaptophysin family of integral

vesicular membrane glycoproteins, which are highly prevalent in synaptic terminals and have been

suggested to assist in SNARE complex formation, SV fusion pore formation, and SSV recycling,

but knockout does not result in any readily apparent synaptic deficits16-18. However, synaptophysin

I and synaptogyrin I double knockout mice have been reported to exhibit normal brain structure,

composition, and SV release probability, but severe deficits in hippocampal short- and long-term

potentiation. Intriguingly, this specific deficit in synaptic plasticity is not observed when other

SVPs are knocked out (e.g. rab3A, synapsin I and II, synaptotagmin I)19. This finding suggests that

synaptophysin-like SVPs are functionally redundant, but are nevertheless essential regulators of

synaptic plasticity, although studies to date have been unable to elucidate the underlying

molecular mechanism.

It also seems that the composition of SV release machinery differs amongst individual SVs, as

well as between SSVs and LDCVs. For instance, synaptophysin and synapsin I have been found

to be enriched in neuronal SSVs, but present at much lower concentrations, or even completely

absent from neuronal LDCVs20,21. Furthermore, studies on the protein composition of single SVs

isolated from rat brains suggest that certain SVPs are sorted with high precision (e.g.

synaptotagmin I), whereas other show significant variability (e.g. synaptophysin, synaptobrevin

2)22. Such differences are likely to have functional consequences for SV trafficking and release

probabilities at individual presynaptic terminals, although this has not yet been directly examined.

A functional correlate of this type of SVP variation is evident in the differential localization and release characteristics of classical and peptide co-transmitters. SSVs are generally found to be directly docked at the AZ and physically linked to VGCCs, whereupon fusion can be triggered by a small Ca2+ microdomain that arises during a single action potential, whereas LDCVs are more

peripherally associated with VGCCs and the AZ, and thus require larger Ca2+ microdomains and

burst firing23,24. This activity-dependent use of distinct co-transmitter substances is one of the

mechanisms by which activity-dependent synaptic plasticity is encoded.

The above observations suggest that the mechanisms regulating the synaptic localization and

release of classical and peptide transmitters are fundamentally different, however, experimental

support for this notion has been difficult to obtain. In large part, this can be accounted for by the

fact that there are few experimental models available where peptidergic synaptic transmission

between pre- and postsynaptic neurons can be examined directly. Indeed, most insights on

neuropeptide release characteristics have so far been derived from non-neuronal, non-peptidergic

(e.g. adrenal chromaffin cells) or non-synaptic (e.g. neurosecretory structures such as the pituitary) models, which do not permit questions abut the nature of synaptic neuropeptide regulation to be addressed. As peptide neurotransmitters are important modulatory signals for synaptic plasticity in many, if not all neuronal circuits25, a fundamental question that remains to be answered is:

what are the intrinsic and extrinsic factors governing peptidergic synaptic transmission, and how

does this influence the function of neuronal networks?

1.2.3 Chemical synapses: specificity of transmission at the postsynaptic terminal

At the PSD, specific neurotransmitter receptors are clustered by physical interactions with

dedicated scaffolding molecules that organize the postsynaptic signaling complex and anchor it to

the cytoskeleton. It is the density at which receptor clusters form and the identity of the receptors

within these clusters that determine the fidelity of synaptic transmission and the characteristics of the postsynaptic response1. Because of its large size and experimental accessibility, the best

understood synapse is that of the neuromuscular junction (NMJ). Here, synaptic transmission

between the presynaptic motoneuron and the postsynaptic muscle fiber is mediated by a single

type of neurotransmitter, and a limited class of neurotransmitter receptors. For instance, synaptic

transmission at the vertebrate NMJ is mediated by ACh, and glutamate at the Drosophila NMJ.

Both of these neurotransmitters and their respective receptors are also widely employed by neurons

of the central nervous system (CNS). Neurons of the CNS, however, face the additional challenge

of receiving innervation from a variety of presynaptic inputs using diverse types of

neurotransmitters, and must meet the increased complexity of synaptic circuitry with mechanisms

for postsynaptic specificity that sort multiple types of neurotransmitter receptors to appropriate

postsynaptic sites. However, some of these scaffolding molecules remain unidentified.

The molecular components of the glutamatergic postsynaptic scaffold, both at the NMJ of

Drosophila and in the CNS of mammals, have been well-defined and appear to be well-conserved as well. For instance, the postsynaptic clustering of glutamate receptors, both at the NMJ and in the CNS, is known to rely upon interactions of the cytoplasmic C-termini of AMPA- and NMDA- type receptors either directly with the PDZ domains of PSD-95/Dlg family scaffolds molecules, or

indirectly via PDZ domain-containing adaptor molecules26-30. The molecular components of the

vertebrate NMJ are similarly well-defined – rapsyn is the molecular scaffold required for muscle

nAChR clustering, although this appears to occur in the absence of PDZ domain interactions31,32.

However, while rapsyn is expressed in neurons, it is not essential for neuronal nAChR clustering.

In mouse ganglionic neurons, rapsyn has been found to be absent from nAChR clusters, and

synaptic nAChR clusters form normally when rapsyn is knocked out33. This finding indicates that

a distinct, as yet undefined molecular scaffold mediates the clustering of neuronal nAChRs. As

postsynaptic receptor clustering is a key determinant of synaptic function, and cholinergic

transmission in the CNS is critical for the induction of synaptic network plasticity underlying learning and memory34, it is imperative to identify these unknown elements of postsynaptic scaffolds to further our understanding of the function, formation and maintenance of specific types of synapses within the brain.

1.3 Synapse Formation

1.3.1 Intrinsic signaling: the ‘ready-set-go’ model of synaptogenesis

At the distal ends of extending neurites, motile structures termed growth cones navigate the extracellular molecular environment and mediate the rapid initial phase of synapse formation via pre-assembled ‘packets’ of synaptic machinery that are established prior to target cell contact. In a ‘presynaptic’ capacity, the growth cones of embryonic Xenopus motoneurons and chick ciliary ganglion neurons are capable of spontaneous and evoked acetylcholine (ACh) release both prior to and immediately after target cell contact35,36. In a ‘postsynaptic’ capacity, embryonic Xenopus myocytes respond to ACh release from motoneurons with nAChR-mediated depolarizations

resembling miniature endplate potentials within minutes of initial contact37. Exogenous

application of serotonin (5-HT) has also been shown to collapse growth cones extending from

isolated Helisoma neurons in a cell type-specific manner, illustrating the influence of early

neurotransmitter-receptor interactions on neuronal architecture and connectivity patterns38. The

structural changes to pre- and postsynaptic growth cones by neurotransmitter-receptor interactions

have since been found to involve Ca2+-dependent remodeling of the actin cytoskeleton39. These

studies demonstrate that prior to target cell contact, an intrinsic genetic program establishes

expression of the minimal elements of the synaptic machinery necessary for the release and

detection of neurotransmitters, as well as the recognition of appropriate synaptic partners. This has

been deemed the ‘ready-set-go’ model of synaptogenesis40.

While growth cones endowed with such synaptogenic packets are able to form functional synapses

initially, these minimal complements of synaptic machinery cannot maintain functional synapses

or promote synaptic consolidation. Previous work from our lab has demonstrated that isolated

growth cones from Lymnaea neurons, termed ‘growth balls’, from pre- and postsynaptic partners

have the innate capacity to form appropriate synapses. Growth ball synapses exhibit functional synaptic transmission within minutes of initial contact, but these preliminary synapses are soon lost without the soma of both neurons41. This finding indicates that synaptogenic factors which are

extrinsic to growth cones must be received at nascent pre- and postsynaptic sites in order for

synaptogenesis to proceed via synaptic differentiation and consolidation.

1.3.2 Cell-cell signaling: cell adhesion molecules induce synaptic differentiation

Following ‘ready-set-go’ priming, bidirectional signaling mediated by cell adhesion molecule

(CAM) interactions trigger the differentiation of pre- and postsynaptic structures through the

recruitment of preformed protein complexes to establish the CAZ and the PSD. Perhaps the most

conspicuous example of CAMs’ synaptogenic role is trans-synaptic neuroligin-neurexin

interactions, which bind the PDZ-domain motifs that underlie the majority of cytoplasmic scaffold

molecule interactions42,43. With heterologous vertebrate cell cultures, neuroligin has been found to

be sufficient to trigger the morphological and functional specialization of the CAZ44, and neurexin

has been shown to complimentarily trigger the differentiation of the PSD45, independent of other

pre- or postsynaptic membrane associated factors. The expression and localization patterns of particular CAM isoforms is an important mechanism for synaptic specificity during

synaptogenesis, as they are known to differentially instruct the formation of distinct types of

synapses. For instance, overexpression of neuroligin-1 or neuroligin-2 has been reported to selectively enhance the formation of excitatory glutamatergic or inhibitory GABAergic synapses, respectively46.

On the presynaptic side, the binding of neuroligin to neurexin establishes a nucleation site for the

assembly of the presynaptic cytoplasmic scaffold47. Highly motile dense-core transport vesicles,

termed Piccolo-Bassoon transport vesicles (PTVs), carry a comprehensive set of proteins required

for the formation of the CAZ: the cytoskeletal scaffold molecules Piccolo and Bassoon; the

presynaptic membrane proteins necessary for regulated SV exocytosis, including the t-SNAREs

Syntaxin and SNAP-25; accessory proteins Munc13, Munc18, Rab3 and RIM; and N-type calcium

channels48,49. Furthermore, it has been demonstrated that PTVs in hippocampal neurons exhibit variable molecular composition, and that fusion of just two precursor vesicles is sufficient to form new presynaptic specializations in a rapid, nearly ‘all-or-none’ type process49. In the search for the

cellular and molecular mechanisms of synaptic specificity, it is therefore an intriguing possibility

that individual presynaptic terminals might be functionally specified during differentiation of the

CAZ as a consequence of the molecular composition of their formative PTVs.

On the postsynaptic side, stationary membrane specializations containing neuroligins serve as

predetermined ‘hotspots’ for synapse formation. During vertebrate glutamatergic synaptogenesis,

for example, clusters of postsynaptic complexes containing neuroligin-1, PSD-95, guanylate

kinase-associated protein (GKAP), and Shank/Homer can recruit PTVs to induce functional

specialization of the presynaptic terminal upon axon contact. Neurexin binding to neuroligin-1

also mediates the subsequent clustering of AMPA-type receptors and recruitment of additional

PSD-95-GKAP-Shank scaffolds to promote further differentiation and maturation of the

PSD27,50,51. In contrast to presynaptic machinery which is delivered to nascent terminals by discrete

packets in mobile transport vesicles, acquisition of postsynaptic machinery, such as AMPA and

NMDA receptors, during the initial stages of postsynaptic differentiation appears to occur

gradually by recruitment from cytoplasmic or non-synaptic surface stocks in the absence of

modular transport52.

Taken together, the above studies demonstrate that genetically predetermined packets of synaptic machinery facilitate the early stages of synapse formation by inducing the differentiation of pre-

and postsynaptic membrane specializations. However, there have been a vast number of

conflicting reports that presynaptic differentiation precedes postsynaptic differentiation, and vice- versa50,53-56. These differing reports seem to suggest that the inductive capacity of these signals in

triggering the initial formation and maturation of new synapses is highly variable. Perhaps this

inherent versatility emerges as a result of discrepancies amongst the cellular, spatial, temporal, or

environmental synaptogenic context, and represent a potential mechanism by which the specific

formation of synaptic connections can be achieved.

1.3.3 Cell-cell signaling: neurotransmitter-receptor interactions

While the precise nature of the intercellular or environmental signals responsible for specificity

during synaptic differentiation have not yet been clearly defined, activity-dependent synaptic

transmission is well known to mediate the phase of synaptic stabilization, refinement, and

maturation that follows the initial establishment of target connectivity and preliminary

differentiation of nascent synapses. This has been elegantly demonstrated by studies of Munc18

(STXBP1) null mice, where neurons assemble morphologically normal networks with differentiated pre- and postsynaptic structures in the absence of neurotransmitter release. However,

synaptic ultrastructure later degenerates and neurons enter apoptosis, revealing that these

rudimentary networks, and the neurons themselves, cannot persist without synaptic transmission57.

The guidance decisions of growth cones and CAM interactions that facilitate the initial structural

organization of the brain, contact between prospective pre- and postsynaptic partners, and the

reciprocal induction of synaptic specializations, are in and of themselves not sufficient to

orchestrate the precise patterns of connectivity required for high-fidelity functions. In a series of

now classical experiments that defined the contribution of synaptic transmission to the remodeling

of developing networks, both spontaneous electrical activity and experience-dependent synaptic

transmission were shown to shape the anatomical and functional organization of retinal projections

and ocular dominance columns of the visual cortex. These events evoke Hebbian plasticity

mechanisms (the strengthening of appropriate synapses with coincident presynaptic action

potentials and postsynaptic ionotropic receptor activation, and the pruning of inappropriate

synapses with asynchronous activity) necessary for activity-dependent synaptic network refinement and the development of visual acuity58,59. Synaptic specificity, then, both at the level

of neuronal network connectivity and the level of the individual synapse, might also be predicted

to be influenced by a variety of factors that determine presynaptic SV release characteristics (e.g.

distinct SVP isoforms) or the profile of receptors that are clustered postsynaptically (e.g. cell-type

specific expression of receptors).

Whereas the influence of synaptic transmission mediated by ionotropic receptors, particularly AMPA- and NMDA-type glutamate receptors, in synapse formation and refinement is well recognized60,

considerably less attention has been paid to the role of metabotropic receptors in synaptogenesis.

Nevertheless, a critical role for GPCR signaling in synapse stabilization and refinement is also

beginning to emerge. The metabotropic glutamate receptor mGluR5 has been best studied in this

context. For instance, activation of phospholipase C (PLC) by mGluR5 has been shown to be essential

for activity-dependent refinement of the somatosensory whisker barrel cortex cytoarchitecture in

mice61,62. mGluR5 activation has also been found to influence the balance of AMPA- and NMDA-type

receptors during excitatory synapse formation63, and mediate the stabilization of glycine receptors during inhibitory synapse formation64. As all neurotransmitters signal through at least one type of

metabotropic receptor, these findings are indicative of a larger and still mostly unexplored function

for metabolic signaling pathways in the refinement of developing synaptic networks by synaptic

transmission.

1.3.3.1 Cell-cell signaling: ‘black-box’ factors – peptide neurotransmitter-receptor interactions

Despite the importance of small molecule classical transmitters to synaptic development, the

contributions of peptide neurotransmitters to synapse formation and maturation remain largely

undefined. This is surprising considering that the signaling of neuropeptides via metabotropic

GPCRs and various second messenger systems have the inherent capacity to modulate a wide variety

of molecular signaling cascades which are likely to influence the trajectory of synaptic differentiation,

stabilization, maturation and refinement. A multitude of studies have established that neuropeptides

are expressed prior to target cell contact, and localized in such a way that would seem to indicate

a similar role in synaptic orchestration. For example, peptidergic vesicles in Drosophila and

Aplysia neurons are rapidly recruited into growth cones and presynaptic terminals by the same

activity-dependent mechanisms that direct synaptogenesis65-67. Other studies of nervous system

development in Lymnaea have demonstrated that peptidergic neurons are some of the first elements

of the nervous system to emerge68, and that some neurons exhibit only transient neuropeptide expression69. Together, these observations are indicative of a novel role for neuropeptides in the

emergence of synaptic specificity during the initial establishment and/or activity-dependent refinement

of synaptic networks.

1.3.4 Extrinsic signaling: neurotrophic factors

Neurotrophic factors (NTFs) are target-derived proteins, often secreted in response to synaptic

activity, that signal through receptor tyrosine kinases (RTKs) to promote neuron survival and

facilitate synapse formation and maturation both pre- and postsynaptically. RTK signaling is

transduced primarily by three downstream signaling cascades: PLC, Phosphoinositide 3-kinase

(PI3K), and mitogen-activated protein kinase (MAPK)70. The differential expression patterns of

various NTF ligands and receptors in the brain, and their distinct region- and cell type-specific actions in neuronal survival and synaptogenesis, are indicative of a system for complex functional specificity which remains the subject of ongoing investigation71.

Of the various types of NTFs, the specific actions of the neurotrophin family in the regulation of

neuronal circuit development and function are currently the best understood. For instance,

neurotrophin-3 (NT-3) and brain-derived neurotrophic factor (BDNF), but not nerve growth factor

(NGF), have been shown to rapidly potentiate developing Xenopus NMJ synapses by promoting

the functional development and maturation of presynaptic terminals72. This effect was later found

to be dependent upon TrkB receptor signaling and MAPK-dependent phosphorylation of synapsin

I73, which mediates SV accumulation and functional docking at AZs14, as well as cytoskeletal

stabilization via actin polymerization mediated by the Rho family GTPase Cdc42, a mechanism

that has also been found to be responsible for the conversion of non-functional ‘silent’ presynaptic

terminals into functional ones74. In mouse spiral ganglion neurons, it has also been reported that

NT-3 promotes whereas BDNF inhibits the accumulation of SNAP-25 and synaptophysin75,

although the actions of these SVPs in driving presynaptic maturation are less clear. While

neurotrophins are known to play specific roles in modulating the postsynaptic function of mature

circuits, for instance, via TrkB-dependent phosphorylation of AMPA and NMDA receptors leading to increased surface expression and conductance, their synaptogenic contributions to postsynaptic development and maturation have been less well-defined, and the underlying mechanisms require further study71.

Perhaps the best characterized role for RTK-mediated signaling in postsynaptic development is

the differentiation and maturation of the vertebrate NMJ induced by the muscle specific kinase

(MuSK) RTK. Release of the proteoglycan agrin from approaching motoneuron terminals binds

to the lipoprotein receptor-related protein 4 (LRP4), which in turn activates MuSK to induce the

clustering of nAChRs and the maturation of postjunctional specializations via the intracellular

effector rapsyn. A phosphorylation cascade initiated by MuSK and effected by Abl and Src family

kinases results in the phosphorylation of rapsyn as well as nAChR β and δ subunits. This

posttranslational modification (PTM) is required for rapsyn-nAChR interactions and their linkage

with the PSD cytoskeleton. Ultimately, postsynaptic differentiation results in the redistribution of

diffusely localized nAChRs (~1,000/μm) to high-density clusters (~10,000/μm) at nerve contact

sites, and a corresponding reduction of nAChR density at extrasynaptic sites (~10/μm)76-78.

The molecular actions of NTFs at individual synapses are furthermore known to be a critical

mediator of the activity-dependent refinement of synaptic circuits. This concept, initially

postulated as the ‘neurotrophic factor hypothesis’, was based on early experimental observations

of the role of neurotrophins in sympathetic ganglia and at the NMJ – competition between co-

innervating presynaptic terminals for a limited amount of NTFs could account for the stabilization

of active presynaptic terminals, and the elimination of inactive ones. For example, Xenopus co-

culture experiments have shown that when a single myocyte is innervated by two motoneurons,

activity-induced secretion of BDNF acts heterosynaptically to stabilize the active presynaptic

terminal and suppress the inactive one79. A similar role for BDNF in activity-dependent refinement

of the mouse visual cortex has also been demonstrated, where overexpression of BDNF was found

to be sufficient to promote the development of ocular dominance columns in the absence of sensory

experience80. This effect might be attributable to BDNF’s stimulation of sodium channel- dependent bursting via TrkB receptors81,82, and may thereby be coupled to synaptic transmission.

1.3.5 Extrinsic regulation of intrinsic signaling: genetic foundations of synaptogenesis

The primary mechanism by which electrical activity, synaptic transmission, and NTF-RTK

signaling converge to direct synaptic maturation and refinement is through regulated gene expression. In addition to the rearrangement of pre-existing synaptic machinery induced by PTMs,

virtually all long-lasting changes in brain function require signaling events between the synapse

and the nucleus that induce transcriptional regulation mechanisms leading to gene activation and new protein synthesis. The appropriate expression of activity-regulated genes underlying the proper assembly and refinement of neuronal circuits requires precise spatiotemporal and stimulus- specific control, and the complexities of the signal transduction cascades that transfer synaptic information to the nucleus and activate gene expression are only beginning to be unraveled83.

1.3.5.1 Activity-dependent gene expression

Ca2+ ions are the key second messengers that link both electrical activity and synaptic transmission to intracellular molecular effectors. Patterned action potentials, such as the depolarizing bursts induced by BDNF82, signal to the nucleus via the activation of L-type VGCCs, which are coupled to a variety of effector proteins such as protein kinase A (PKA) anchoring protein (AKAP), Src tyrosine kinases, or Ca2+/calmodulin-dependent protein kinase II (CaMKII). Nuclear signaling via synaptic activity can be achieved by a multitude of mechanisms: (i) the activation of Ca2+-fluxing ionotropic receptors, such as nAChR or NMDA; (ii) the activation of L-type VGCCs in response to membrane depolarizations induced by non-Ca2+-fluxing ionotropic receptors, such as AMPA; or (iii) the release of Ca2+ from internal stores through IP3 receptors in response to the activation of metabotropic receptors that signal via PLC. The precise route of Ca2+ entry and pattern of intracellular Ca2+ oscillations are also known to be important for determining the specificity of gene induction in response to activity, influenced by various channel properties such as conductance, open time, subcellular localization, and coupling to effector proteins83,84.

The first indication that neuronal and synaptic activity might influence the expression of specific

genes came from the observation that both membrane depolarization and neurotransmitters could

trigger rapid, transient, and selective transcriptional programs in rat PC12 cells, which differentiate

into sympathetic neurons that express nAChR after prolonged NGF treatment. Exposure of

differentiated PC12 neurons to the nAChR agonist nicotine was found to rapidly, transiently, and

specifically increase transcription of the c-fos proto-oncogene and β-actin, dependent upon the

activation of VGCCs. Another intriguing finding from this study which highlights the convergence

of synaptic maturation mechanisms was that NGF could similarly induce gene-specific

upregulation, but through a slower, Ca2+-independent mechanism85. Over 300 genes are now

known to exhibit rapid and transient expression in response to neuronal activity, and recent

progress in this area has lead to the identification of a number of synapse-associated proteins (e.g.

synapsin I86) and neurotrophic factors (e.g. BDNF83) as activity-induced genes. It is of interest to note that the molecular mediators of activity-dependent transcription necessary for synaptic maturation and refinement also exhibit a striking degree of overlap with the causative genetic mutations of several human cognitive disorders which emerge after an apparently typical period of early postnatal development, such as the Autism spectrum disorder Rett syndrome (MECP2)83.

Taken together, these observations indicate that precise patterns of activity induced by

neurotransmitter-receptor and NTF-RTK interactions stimulate the expression of specific gene

products that have the capacity to directly modulate synaptic function. However, given the

complex nature of gene regulation and transcription networks, is likely that there are a number of

activity-induced gene products involved in synaptic maturation and refinement whose influence

remains unrecognized. Transcriptional regulatory mechanisms such as epigenetic modification or

cell type-specific activity of promoter elements would furthermore be uniquely suited to modulating the genetic foundations of synaptic specificity.

1.3.5.2 NTF-dependent gene expression

Target-derived NTFs are also known to mediate gene-specific transcriptional upregulation through

activity-independent mechanisms. This can most likely be attributed to RTK-dependent activation

of MAPK cascades leading to the phosphorylation of transcription factors, although the precise

molecular cascades and their gene targets, particularly in neurons, remain to be fully elucidated.

At the vertebrate NMJ, motoneuron-derived neuregulin (initially called ARIA, for AChR inducing

activity) signaling via muscle-derived ErbB RTK is known to induce postsynaptic maturation via

the subunit-specific transcription of nAChR δ and ε genes, activated by a multisubunit Ets family

transcription factor87-89. In peripheral ganglionic neurons and central hippocampal neurons, neuregulin has been shown to selectively upregulate the transcription of neuronal nAChR α3 and

α7 subunits, demonstrating that the functional role for neuregulin-ErbB signaling in regulating

nAChR subunit composition during cholinergic synaptogenesis is conserved between muscles and

neurons90-92. Neuregulin has also been found to induce the expression of NMDA receptor subunit

NR2C during synaptic maturation in the cerebellum93. Furthermore, BDNF has been reported to

upregulate the expression of AMPA receptor subunits GluR1-3 in cultured hippocampal neurons,

although whether this occurs in an activity-dependent or -independent manner remains to be

determined94. Together, these observations outline a well-documented and seemingly widely- applicable, but not yet fully realized model for activity-independent transcription in postsynaptic maturation: target-derived NTFs influence the subunit composition, and thereby tune the function,

of neurotransmitter receptors. However, in many instances, the precise identity of the transcription

factors that link NTF-RTK signaling to the expression of mRNA for specific neurotransmitter

receptor subunits remain unidentified.

1.3.5.3 Signal integration: coincidence detection

The rapid expression of activity-induced genes, many of them transcriptional regulators (e.g. c- fos), required for synapse formation and long-term forms of facilitative synaptic plasticity almost always involve phosphorylation of the transcription factor CREB (cAMP response element [CRE]

95 binding protein) . The canonical CREB cascade is effected by Gαs-coupled GPCR signal

transduction: (i) adenylate cyclase (AC) stimulation; (ii) cAMP synthesis; (iii) PKA activation;

(iv) CREB phosphorylation; and (iv) transcriptional activation of genes with CRE promoter

elements. Studies in rat PC12 cells have demonstrated that CREB is phosphorylated by CaMKII

downstream of L-type VGCCs96, and the MAPK pathway downstream of RTK97, indicating that

CREB acts as a molecular ‘hub’ to coordinate and appropriately influence gene expression in

response to impulse activity, synaptic transmission, and NTF signaling. Moreover, simultaneous

phosphorylation of CREB by PKA and CaMKII at distinct sites has been shown to potentiate c-

fos expression above the inductive effects of PKA or CaMKII alone98. The convergence of discrete

intracellular second messenger pathways on CREB, and likely other transcription factors as well,

functions as a synergistic mechanism for coincidence detection: stimulus-specific additive effects promote synapse maturation by bolstering transcription-dependent remodeling, whereas sub- threshold effects maintain nascent synapses in an immature state that may promote elimination.

1.3.5.4 Signal integration: ‘black-box’ factors – the MEN1 gene

While the functions of some activity-induced genes, such as BDNF or synapsin I, in synapse

formation and maturation are readily apparent, the precise roles of others are less conspicuous.

Previous work from our lab lead to the discovery that menin, the protein product of the MEN1 (for

Multiple Endocrine Neoplasia type 1 syndrome) tumor suppressor gene, is a synaptogenic factor.

We have previously demonstrated that: (i) postsynaptic MEN1 expression is necessary and sufficient for cholinergic synaptogenesis between identified Lymnaea neurons; (ii) MEN1 expression is induced by NTFs, specifically EGF acting at postsynaptic EGF RTKs (EGFR); (iii) both EGF/R and MEN1 induce the functional expression of nAChR; and (iv) these effects require

EGF/R-induced bursting activity, L-type VGCC and MAPK signal transduction99-102. Thus, it

appears that in Lymnaea CNS neurons, activity- and NTF-dependent signaling via L-type VGCCs and MAPK converge on MEN1 gene induction to promote postsynaptic cholinergic development via menin. Furthermore, subsequent studies by other groups in both mouse and rat models of peripheral nerve injury (which induces NTF expression, neurite sprouting, and synaptic remodeling in the spinal cord dorsal horn103,104) have shown that the upregulation of MEN1 mediates synaptic plasticity and the development of neuropathic pain105-107. These observations

indicate that menin’s synaptogenic function has been conserved across evolution; what remains

to be determined, however, are the underlying mechanisms.

In general, tumor suppressor gene functions are directly dependent upon trophic factor stimulation,

and a wealth of tumor suppressors have been implicated in various aspects of synaptic regulation,

both in the nucleus and directly at the synapse26,108-110. As a tumor suppressor, the function of

menin in cancer biology, particularly with regards to transcriptional regulation, has been well

studied. Menin is known to act as a nuclear scaffold that binds to cell signaling proteins (e.g. β-

Catenin) as well as transcriptional activators (e.g. MLL1 [mixed-lineage leukemia 1]) and

repressors (e.g. JunD) to mediate stimulus- and gene-specific transcription necessary for cell cycle regulation111. Conversely, the molecular actions of menin in neurons have never before been

investigated. All together, a logical hypothesis to be derived from the MEN1 cancer literature, our observations regarding the role of EGF/R and menin in Lymnaea neurons, and the regulation of nAChR by EGF/R-related neuregulin-ErbB (see above), is that menin induces gene-specific transcription of nAChR subunits to tune the function of neuronal nAChRs and promote postsynaptic maturation in response to EGF.

1.4 Synaptic Specificity

1.4.1 Cell-cell signaling: the ‘chemoaffinity’ model of synaptic specificity

As the proper anatomical and structural wiring of brain circuits occurs in the absence of synaptic

transmission, the mechanisms that instruct target specificity must be encoded by a genetically predetermined program prior to synapse formation. In the most classical example of synaptic specificity, the ‘chemoaffinity’ hypothesis posits that unique CAM signatures facilitate the matching of presynaptic terminals to appropriate postsynaptic sites, such that target recognition and synaptic specificity arises as the result of homophilic interactions. Several families of CAMs

(e.g. immunoglobulin, cadherin) are encoded by multiple genes and/or by complex singular genomic loci that produce large numbers of distinct protein isoforms through alternative RNA splicing. The gene for Drosophila Dscam, for example, encodes over 19,000 different isoforms of the immunoglobulin-type CAM, and a single neuron is estimated to express between 10-50 isoform variants112. These diverse trans-synaptic adhesion proteins mediate specification of synaptic connectivity by the facilitation of cellular target recognition and by the differentiation of individual synapses113.

The hippocampus is a widely-used model for the study of synaptic specificity because different

cell types exhibit highly specific patterned connectivity with structurally distinct classes of

synapses (e.g. dentate gyrus [DG]→CA3 mossy fiber). While synaptic specificity was generally

assumed to be lost in dissociated cell culture approaches used for the in vitro study of

synaptogenesis, mouse DG neurons have recently been shown to preferentially innervate their

endogenous CA3 targets and form mossy fiber-type synapses. The recapitulation of circuit specificity in vitro was found to be mediated by the cadherin-9 isoform, which is exclusively expressed by both DG and CA3 neurons of the hippocampus114. This observation demonstrates

that intrinsic, cell type-specific genetic programs regulate the expression of CAM isoforms to

promote appropriate interactions between distinct types of neurons, leading to the differentiation

of particular types of synapses. However, as in synaptogenesis, the specification of cellular and

subcellular target selection by a remarkably diverse repertoire of CAMs is not sufficient to

establish appropriately functioning neuronal networks.

1.4.2 Signal integration: mechanisms that specify the function of individual synapses

After guidance decisions and CAM interactions facilitate target selection and synaptic

differentiation, synapse-specific maturation proceeds via NTF- and activity-dependent

mechanisms that act to refine both network connectivity and the function of individual synapses.

Indeed, it is now generally accepted that the molecular composition of mature synapses is

continually modified by bidirectional signaling between pre- and postsynaptic partners and by

signals from the environment, such that the functional specificity of each individual synapse

ultimately reflects the history of activity and molecular signaling events that influenced its

differentiation, maturation, and plastic remodeling115.

1.4.2.1 Cell-cell signaling: NTF-dependent interactions

A balance between excitatory and inhibitory innervation is essential to the proper function of the

brain, and an imbalance of specific types of synapses often results in neurological disorders such

as epilepsy. Two variants of fibroblast growth factor (FGF), acting at distinct receptors, have been

shown to act as region-specific target-derived organizers of glutamatergic and GABAergic

presynaptic terminals innervating pyramidal neurons of the hippocampal CA3 region. Specifically,

FGF22 null mice were found to exhibit a defect of SV clustering at CA3 glutamatergic terminals,

whereas clustering defects were observed at CA3 GABAergic terminals in FGF7 null mice.

Localization of the presynaptic AZ protein Bassoon, or the postsynaptic scaffolds PSD95 or

gephyrin (the molecular scaffold for GABA receptors), however, were not affected in either

mouse, suggesting that differentiation and maturation processes are independent, and that FGF

isoforms play a selective role in presynaptic maturation. Dendritic FGF22 puncta exhibited co-

localization with PSD95, and FGF7 exhibited co-localization with gephyrin, indicating that these

NTFs are targeted to distinct dendritic subdomains to promote synapse-specific development116.

These observations suggest that distinct NTFs control the maturation of specific types of synapses

in particular brain regions, and thereby modulate the function of neuronal networks. This concept

could conceivably be applied to the functional specificity of individual synaptic terminals as well,

such that spatially segregated presynaptic terminals of an axon might experience dissimilar

environments of target-derived NTFs, which may differentially influence presynaptic maturation,

molecular composition and SV release characteristics.

1.4.2.2 Cell-cell signaling: activity- and neurotransmitter-receptor dependent interactions

One of the mechanisms underlying the induction of long-term plasticity in the mature nervous

system is the formation of new synapses. This process is influenced by the cell- and synapse- specific actions of neurotransmitters, signaling through GPCRs and intracellular second

messengers, to modulate the surface localization and gene expression of CAMs. For example, the

Aplysia gill-withdrawal reflex is a simple network consisting of a monosynaptic connection

between a sensory neuron (SN) and a motoneuron (MN), which is modulated by serotonergic and

FMRFamidergic interneurons that synapse onto the SN. In vitro recapitulated SN-MN synapses exhibit long-term facilitation (LTF) in response to repeated applications of 5-HT, and long-term depression (LTD) in response to repeated applications of FMRFamide, mediated by changes in presynaptic (SN) connectivity. LTF is dependent upon signal transduction to the nucleus via the cAMP-PKA-CREB cascade, which results in transcriptional down-regulation of immunoglobulin- type apCAMs. 5-HT also acts via local cAMP-PKA signaling to induce the rapid endocytosis of extrasynaptic apCAM, which evokes SN neurite defasciculation to promote the formation of new

SN-MN connections. Moreover, these changes are counteracted or induced in the opposite direction by FMRFamide signal transduction through a cAMP-independent pathway, and are not paralleled in the postsynaptic (MN) neuron117-120. These observations suggest that neurotransmitters achieve synapse-specific remodeling as a function of (i) the activity of specific neurons within a synaptic network, and (ii) the specific profiles of signaling cascades induced in particular neurons by particular metabotropic neurotransmitter receptors.

Another mechanism that has been found to underlie this form of synaptic specificity in Aplysia is the activity-dependent trafficking and local synthesis of new synaptic proteins by the ‘molecular tagging’ of activated synapses. Short-term facilitation (STF) of the SN-MN synapse occurs in response to a single application of 5-HT, which induces covalent modification (e.g. phosphorylation by PKA) of existing synaptic machinery that occurs rapidly and transiently. LTF, on the other hand, requires repeated applications of 5-HT, gene transcription, protein synthesis,

and the growth of new synaptic connections, which occurs after a delay and persists for days. In a

culture system where a bifurcated SN forms synapses with two MNs, one on each axon, repeated

application of 5-HT to one synapse has been shown to induce LTF via CREB-dependent

transcription and neurite sprouting at the activated, but not the inactive synapse. When synapse- specific LTF is induced, a single pulse of 5-HT at the opposite synapse is sufficient to ‘tag’ the synapse with a PKA-dependent activation mark for LTF that allows it to capture the products of

CREB-dependent transcription and induce synapse-specific growth. One of the gene targets of

CREB necessary for synapse-specific trafficking was recently identified as kinesin, a molecular motor that transports synaptic cargos such as vesicles, mRNA and new synaptic proteins. It is also of interest to note that this process has been found to require synapse-specific mRNA localization and local protein synthesis of the sensory neuron specific neuropeptide sensorin, which acts as a neurotrophic-like factor to further promote the growth and stabilization of new synapses121-125.

This phenomenon of ‘synaptic capture’ addresses a paradoxical cell biological problem and

identifies a mechanism through which transcription- and protein synthesis-dependent synaptic

plasticity can occur at individual synapses despite the fact that all synapses of a neuron share a

single nucleus.

1.4.2.3 Cell-cell signaling: ‘black-box’ factors – presynaptic transmitter specificity

In addition to the mechanisms that differentiate synaptic function via structural remodeling, the

presynaptic terminals of individual neurons have also been found to exhibit distinct innate

functional characteristics, in terms of molecular composition, release probability126, and the types

of co-transmitter substances that are localized to or released from presynaptic AZs127. This

diversity is thought to be critical for the selective processing and non-uniform relay of information

that is required for functional flexibility within a circuit, although the mechanisms underlying this

phenomenon remain mostly undefined. One of the key determining factors that has been widely

recognized, however, is the influence of postsynaptic identity on the function and molecular

composition of presynaptic terminals128-134. Heterogeneity amongst the presynaptic terminals of a

single neuron is currently best characterized by the selective induction of synaptic plasticity

mechanisms that differentiate vesicle release machinery at synapses with distinct postsynaptic

targets. For instance, tetanic stimulation has been found to induce a presynaptic form of

cAMP/PKA-dependent long-term potentiation (LTP) at hippocampal mossy fiber synapses with

CA3 pyramidal neurons, but cAMP/PKA-independent long-term depression (LTD) at synapses

with inhibitory interneurons128. The synapse-specific induction of LTP and LTD was found to be

dependent upon the selective targeting and regulated surface expression of mGluR7, (the

presynaptic auto-receptor that couples high-frequency stimulation to cAMP/PKA), under the

control of postsynaptic Ca2+ and presynaptic protein kinase C (PKC)134-136. This finding suggests

that the molecular composition of individual presynaptic terminals is differentiated by some

unidentified trans-synaptic signal that is specific to the identity of the postsynaptic target.

A functionally related form of presynaptic specificity in which the presynaptic terminals of invertebrate neurons exhibit differences in either presynaptic neurotransmitter composition or release characteristics has also been suggested to depend on target cell derived signals, but remains mechanistically uncharacterized. For example, studies of in vitro reconstructed synapses between identified Aplysia neurons have demonstrated that a co-transmitting neuron varies the use of classical small molecule transmitters or neuropeptides according to the identity of the postsynaptic

target137,138. Similarly, synaptic transmission between a co-transmitting neuron and two distinct postsynaptic targets within the in situ stomatogastric ganglion network of Cancer has been found to be differentially mediated by the release of classical and peptide neurotransmitters139. In these

instances, the postsynaptic neurons have been shown to respond to the exogenous application of

both classical and peptide neurotransmitters used by the presynaptic neuron, indicating that this

form of synaptic specificity is not due to variations in postsynaptic receptor expression.

Transmitter specificity therefore must arise presynaptically, however, whether this is due to

differential localization or release characteristics of SSVs and LDCVs remains to be determined.

On the one hand, the above examples suggest that presynaptic specificity is an innate property of

individual synaptic terminals that arises from target cell-dependent interactions during presynaptic

maturation. On the other, they also allude to the possibility that functional specificity may be a

plastic property that can be modulated in mature circuits by experience or the local environment.

Synapse-specific differentiation necessitates some form of context-dependent, target-cell specific

retrograde signaling mechanism, however, the molecular identity of this signal remains elusive.

Considering that NTF signaling and neurotransmitter-receptor interactions play important roles in

synapse formation and maturation, these are attractive candidate mechanisms for the functional

differentiation of individual synaptic terminals as well. There are abundant examples of cell-type

specific expression of synapse-associated proteins, including various isoforms of NTFs, RTKs, ionotropic receptor subunits, and metabolic neurotransmitter receptors that exhibit differential G protein coupling115,140,141. An intriguing hypothesis, therefore, is that presynaptic transmitter

specificity arises from the target cell-specific actions of neurotransmitters, acting at distinct receptors to induce unique profiles of retrograde transmitters to differentiate the molecular

composition and function of individual presynaptic terminals, and that this may furthermore be influenced by the local environment of NTF-RTK signals an axon terminal encounters.

1.5 Specific background and rationale of the Lymnaea model system

The use of simple model systems and experimental preparations has contributed to numerous fundamental advancements in our understanding of the nervous system. Although distinctions can be made amongst organisms (e.g. between invertebrates and vertebrates142) or experimental preparations (e.g. between the NMJ and the CNS143), the prevalence of similarities emphasizes the

fact that the basic tenants of the cellular, molecular, and genetic mechanisms governing synapse

formation, specificity, and plasticity have been highly conserved. While the ultimate goal of

studies in neuroscience may be to define the human condition, the complexity of the vertebrate

CNS, with billions of neurons and trillions of synapses, confounds investigations of the formation

and function of individual synapses. The CNS of the invertebrate gastropod mollusc Lymnaea

stagnalis, by contrast, is comprised of only ~20,000 cells, where individual neurons of known

neurotransmitter phenotype can be readily identified by their size, location, and colour.

Furthermore, the contributions of individual neurons and their synaptic connections within limited

networks to the generation of readily observable behaviours such as respiration144,145 or feeding146

have been defined. The phenomenon of synaptic specificity is strikingly exemplified by the

selectivity with which synapses form between identified invertebrate neurons in culture,

recapitulating the same patterns of synaptic connectivity and directionality that are observed in

the intact CNS147-149. This feature makes invertebrate models an ideal preparation for the study of

synaptic specificity. Individual cloning efforts and the recent generation of a CNS transcriptome

library150 have also now provided sufficient sequence information to make the Lymnaea model system amenable to genetic investigation and manipulation. Lymnaea CNS neurons, through both in vitro and in vivo experimentation, furthermore offer the unique opportunity to ascribe the effects

induced by various types of in vitro manipulations to a particular neuron to the function of synaptic networks and the control of behaviour.

1.5.1 Synaptic cell culture: the soma-soma model

One of the constraints on synaptogenesis is the requirement for neurite outgrowth, which depends upon NTF signaling151. To circumvent this, the soma of isolated invertebrate neurons can be

juxtaposed in culture, allowing for synapse formation in the absence of NTFs. This soma-soma

configuration provides a preparation where the cellular, molecular, genetic and temporal

foundations of the synaptogenic program can be selectively dissected in the absence or presence

of NTFs, by the addition of purified NTFs to the culture media or the use of NTF-rich CNS- conditioned media152. Importantly, this experimental resolution cannot be equivalently attained

with mammalian dissociated cell culture approaches, as neurons do not extend processes, form

synapses, or survive without trophic support. Soma-soma paired neurons form pre- and post-

synaptic specializations that exhibit electron-dense appositions, SV and VGCC clustering153-155.

Presynaptic neurons typically form multiple release sites with appropriate synaptic partners, and

parallels can be drawn with the synaptic morphologies of the Calyx of Held- or mossy fiber-type

synapses observed in the mammalian brain. This offers a unique opportunity to define the nature

of molecular signaling interactions between neurons in an amplified context of synaptic

homogeneity. Innervation characteristics also exhibit a remarkable degree of fidelity, as a single

presynaptic neuron cultured with two identical, otherwise appropriate postsynaptic targets will

selectively innervate only one156. The soma-soma model also provides an ideal preparation in

which to investigate the influence of target cell identity on the molecular composition and function

of presynaptic terminals, as neurons can be paired selectively with appropriate or inappropriate partners, as well as simultaneously with multiple distinct partners. A brief overview of the experimental approaches used here with the Lymnaea model system is presented in Appendix I

(Chapter 6).

1.5.2 The Lymnaea model: nicotinic acetylcholine receptors

As described above, previous work from the Syed lab has defined the role of EGF/R signaling and

MEN1 gene induction in the regulation of interneuronal cholinergic synaptogenesis. In contrast to the exclusively cation-conducting (excitatory) nAChR channels of vertebrates, invertebrates express both cation- and anion-conducting (inhibitory) nAChR channels157. Previous work from our lab and collaborators in the Netherlands have determined the sequences of the 12 Lymnaea nAChR subunits, as well as a preliminary characterization of their expression patterns and contributions to the formation of functional excitatory or inhibitory nAChR channels158,159. Both

EGF/R signaling and MEN1 expression induce a functional switch in nAChR expression profiles, from anionic to cationic channels. Whether EGF/R and MEN1 lead to the transcriptional regulation of distinct nAChR subunits, or the redistribution of nAChR surface expression is currently unknown. Although this biphasic cholinergic system is unique to invertebrates, it offers an ideal clear-cut electrophysiological assay with which to investigate NTF- and MEN1-mediated effects on the regulation of nAChR expression and function during cholinergic synaptogenesis. Vertebrate nAChR preparations, by contrast, rely on the interpretation of more subtle changes in channel properties such as ACh sensitivity, Ca2+ permeability, or conductance.

1.5.3 The Lymnaea model: FMRFamide and peptidergic synaptic transmission

The FMRFamide tetrapeptide (Phe-Met-Arg-Phe-NH2), isolated from the mollusc Macrocallista,

was the first peptide substance found to function in synaptic transmission160. The FMRFamide

gene as well as the biological actions and cell-specific expression patterns of FMRFamide-derived neuropeptides have been studied in great detail – particularly in Lymnaea, making it a well-defined system in which to study the role of neuropeptides in synapse formation161-163. Although the

metabotropic FMRFamide receptor is yet to be identified, its cell type-specific functional characteristics have been extensively characterized, as described above9,10. The sequences of

several distinct Lymnaea G proteins have also been characterized, and they are similarly known to

exhibit cell type-specific patterns of expression in the CNS164-166. Furthermore, five genes

encoding FMRFamide-related neuropeptides (FaRPs) have been identified in mammals, whose

neuropeptide products modulate diverse behaviours such as the stress response, reproduction,

nociception, feeding, and cardiovascular regulation167. Therefore, findings on the function of

invertebrate FMRFamide neuropeptides are likely to be translatable, and provide some of the first

mechanistic insights into the regulation of neuropeptides at synapses, particularly with regard to

their influence on synapse formation, synaptic specificity, and the regulation of neuronal networks.

FMRFamide neuropeptides have been shown to inhibit neurotransmitter secretion via GPCR

signaling168,169. Previous work from our lab has demonstrated that this effect regulates the timing

of synaptogenesis between reciprocally connected neurons of the Lymnaea respiratory central

pattern generator (rCPG)170, and that this property can be modulated by AA. Specifically, a

FMRFamidergic neuron was shown to be the first to innervate its non-FMRFamidergic

postsynaptic target, and that the postsynaptic neuron gained secretory capabilities as synaptic

maturation proceeded. This observation indicates that the formation of a reciprocal synapse may

require the loss of FMRFamide-mediated synaptic transmission. Furthermore, this study provides

an important experimental lead for further dissection of the complex phenomenon of presynaptic

transmitter specificity: (i) in the Lymnaea CNS, there may be similar instances of co-transmitting neurons varying the use of classical and peptide neurotransmitters at individual presynaptic terminals, according to the identity of postsynaptic targets; and (ii) that the identity of the retrograde signal that induces synapse-specific differentiation may be the selective production of

AA, according to the target cell-specific coupling of metabotropic GPCRs to distinct G protein complexes.

1.6 Specific Aims

Main Objective: To elucidate the activity- and NTF-dependent mechanisms responsible for establishing pre- and postsynaptic specificity during synaptogenesis (Fig. 1.1).

Specific Aim 1: To determine the molecular mechanisms underlying the synaptogenic function of menin in postsynaptic development.

Specific Rationale and Hypothesis 1: The MEN1/menin tumor suppressor influences synapse formation and synaptic plasticity in the CNS, however, the molecular function of menin in neurons remains unidentified. Previous work from our lab has demonstrated the necessity and sufficiency of postsynaptic MEN1 expression in NTF-dependent excitatory cholinergic synaptogenesis

between identified Lymnaea neurons. As menin is known to be a nuclear protein and

transcriptional regulator, my hypothesis is that menin mediates the selective transcriptional

upregulation of cationic nAChR subunits to promote excitatory cholinergic synaptogenesis in response to NTF signaling.

Specific Aim 2: To determine whether the molecular actions of menin in neurons are conserved across evolution.

Specific Rationale and Hypothesis 2: MEN1 homologues are found throughout the animal kingdom, and have been highly conserved across evolution. The factors regulating specific patterns of neuronal nAChR expression and localization are not well defined. MEN1 expression is

found throughout the mammalian CNS, with especially high levels in the hippocampus, a center

that receives extensive cholinergic innervation and expresses high levels of nAChR. As the

synaptogenic function of menin has been evolutionarily conserved, my hypothesis is that the

synaptogenic actions of menin in the mouse CNS are specifically tied to cholinergic postsynaptic

development via the regulation of nAChR expression and function.

Specific Aim 3: To determine the molecular mechanisms that differentiate the peptidergic

composition and function of individual presynaptic terminals.

Specific Rationale and Hypothesis 3: Neuropeptides are common co-transmitter substances

important for synaptic plasticity, but their roles in synaptogenesis and the mechanisms regulating

their synaptic release are largely unknown. Several lines of evidence indicate that the transmitter

composition and release characteristics of individual presynaptic terminals are influenced by target

cell identity. Exactly how this form of presynaptic specificity is established remains unresolved.

To address these questions, I will investigate the peptidergic characteristics of synapses between

an identified Lymnaea presynaptic neuron and two distinct postsynaptic targets. My hypothesis is

that trans-synaptic molecular signaling interactions induced by target cell-specific neuropeptide- receptor interactions differentiate the neurotransmitter release characteristics of individual presynaptic terminals.

Figure 1.1- Overview of specific aims and hypotheses 40

Figure 1.1: Overview of specific aims and hypotheses

Specific Aim 1: To determine the postsynaptic function of MEN1 in excitatory cholinergic

synaptogenesis between Lymnaea neurons. EGF/R signaling stimulates the MAPK cascade and L-

type VGCC-dependent bursting in the postsynaptic neuron, these signals converge on MEN1 gene

activation101,102. Menin in turn acts as a transcriptional activator of excitatory nAChR subunits to

induce the functional expression of excitatory nAChR channels. EGF/R signaling furthermore

mediates the postsynaptic clustering of excitatory nAChR channels via the phosphorylation of an

unknown molecular scaffold for neuronal nAChR.

Specific Aim 2: To determine the role of MEN1 in regulating the cholinergic postsynaptic function

of mouse hippocampal neurons. Menin acts as a subunit-selective transcriptional activator of

neuronal nAChR subunits, possibly nAChRα58,106, the expression of which influences the

functional properties and/or trafficking of neuronal nAChR channels.

Specific Aim 3: To elucidate the molecular mechanisms that determine the use of classical and

peptide neurotransmitters at individual presynaptic terminals. Invertebrate co-transmitting neurons

selectively release classical or peptide neurotransmitters at synapses with distinct postsynaptic

targets137-139. The FMRFamide GPCR exhibits cell type-specific G protein coupling, producing

postsynaptic excitation or inhibition9,10. FMRFamide excitatory postsynaptic responses stimulate

AA synthesis via Ca2+- and Gβγ-dependent activation of phospholipase A2 (PLA2)171. AA acts

retrogradely at the presynaptic terminal, perhaps via PKC172, to modulate SV localization and/or

the function of SV release machinery to selectively facilitate or inhibit SSV or LDCV release.

Signal transduction to the nucleus may regulate the expression and/or selective trafficking of SVPs and FMRFamide LDCVs to synapses that have been ‘molecularly tagged’ by AA effectors.

Table 1.1 - Cellular and molecular mechanisms underlying synaptic specificity

Determinants of Cellular Actions Representative Mechanism Synaptic Specificity Chemoaffinity CAM-dependent target recognition DG→CA3 mossy fiber synapse ‘molecular formation via cell type-specific matching’ expression of Cadherin-9114 Distinct actions of Selective promotion of synaptic Postsynaptic FGF22 and FGF7 NTF-RTKs maturation mediate the maturation of glutamatergic and GABAergic presynaptic terminals116 Neurotransmitters Variable release characteristics at Selective release of classical or individual presynaptic terminals of a peptide neurotransmitters at single neuron synapses with two distinct postsynaptic targets137-139 Mechanism unknown Neurotransmitter Variations in postsynaptic response Cell type-specific excitation or receptors (e.g. distinct metabotropic signaling inhibition by FMRFamide9,10 cascades leading to excitation or inhibition; ionotropic channel nAChRα5 subunit modulates conductance, Ca2+ permeability, receptor channel properties6,7 modulated by subunit composition)

Segregation of receptors to Dedicated scaffold molecules appropriate postsynaptic sites (e.g. PSD-95, gephyrin) SVPs SV localization and release Ca2+ sensitivity of release characteristics influenced by the determined by synaptotagmin expression of selective SVP variants (e.g. 1, IV, 7)12,15,173,174 isoforms or the quantity of SVPs in the SV Synaptophysin in SSVs and LDCVs is highly variable21,22 Mechanism unknown Gene expression Selective ‘capture’ of the gene 5-HT-cAMP-PKA-CREB- products for synaptic differentiation dependent transcription, or plasticity by ‘molecular tagging’ kinesin-dependent trafficking of of activated synapses new synaptic machinery125

Activity-induced genes, expression Expression of synapsin I or of SVPs influenced by electrical synaptotagmin IV induced by activity or synaptic transmission activity86,175 Retrograde signaling Influence of postsynaptic identity on Mechanism unknown the molecular composition and (postsynaptic Ca2+ signaling function of individual presynaptic and presynaptic PKC?)134,135 terminals (selective AA synthesis?)170

Table 1.2 - A partial list of presynaptic proteins and their functions

Synaptic Vesicle Proteins Function Integral Membrane Proteins Synaptotagmin (Syt) Ca2+ sensor for vesicle release Dual function, facilitative (Syt I in mammals; Syt I C2B-β in mollusks) or inhibitory (Syt IV; Syt I C2B-α) Synaptophysin family (Syp) Primary function unknown (synaptophysin, synaptoporin, Synaptobrevin orientation? Fusion pore formation? synaptogyrin) Synaptic vesicle recycling? Synaptobrevin 2 (Syb) Membrane fusion, vesicular (v)-SNARE (aka VAMP2) (soluble N-ethylmaleimide sensitive fusion protein attachment protein receptor) Synaptic Membrane Proteins Function Integral Membrane Proteins Syntaxin (Syx) Membrane fusion, target (t)-SNARE SNAP-25 Membrane fusion, t-SNARE Vesicle Associated Proteins Function Complexin Clamps the primed SNARE complex Synapsin I Docks vesicles at the active zone Munc 13/18 Vesicle priming, opens/closes Syntaxin VGCC Activity-dependent Ca2+ influx (N-type, P/Q-type) α1 subunit binds syntaxin, SNAP-25, Synaptotagmin Presynaptic Modular Proteins Function Piccolo/Bassoon Presynaptic active zone scaffold (contains PDZ motif) Rab3 Small G proteins, regulates vesicle fusion RIM Presynaptic active zone scaffold (PDZ) (Rab3 interacting molecule) CASK Presynaptic active zone scaffold (PDZ) (Ca2+/Calmodulin-dependent serine protein kinase) CAST Presynaptic active zone scaffold (PDZ) (Cytomatrix associated structural protein) Mint, Veils Presynaptic active zone scaffold (PDZ) Binds CASK, Munc18 Liprin-α Synaptic vesicle/protein recruitment Binds RIM, CAST, CASK, Mint, Munc13

Chapter Two: Two proteolytic fragments of menin coordinate the nuclear transcription and postsynaptic clustering of neurotransmitter receptors during synaptogenesis between

Lymnaea neurons

This chapter has been published in the following manuscript:

Getz A, Visser F, Bell EM, Xu F, Flynn NM, Zaidi W & Syed NI. Two proteolytic fragments of menin coordinate the nuclear transcription and postsynaptic clustering of neurotransmitter receptors during synaptogenesis between Lymnaea neurons. Scientific Reports. 6, 31779; doi 10.1038/srep31779 (2016). – Reproduced with permission from Macmillan Publishers Ltd, Nature Publishing Group.

2.1 Abstract

Synapse formation and plasticity depend on nuclear transcription and site-specific protein targeting, but the molecular mechanisms that coordinate these steps have not been well defined.

The MEN1 tumor suppressor gene, which encodes the protein menin, is known to induce synapse formation and plasticity in the CNS. This synaptogenic function has been conserved across evolution, however the underlying molecular mechanisms remain unidentified. Here, using central neurons from the invertebrate Lymnaea stagnalis, I demonstrate that menin coordinates subunit- specific transcriptional regulation and synaptic clustering of nicotinic acetylcholine receptors

(nAChRs) during neurotrophic factor (NTF)-dependent excitatory synaptogenesis, via two proteolytic fragments generated by calpain cleavage. Whereas menin is largely regarded as a nuclear protein, my data demonstrate a novel cytoplasmic function at central synapses.

Furthermore, this study identifies a novel synaptogenic mechanism in which a single gene product

coordinates the nuclear transcription and postsynaptic targeting of neurotransmitter receptors through distinct molecular functions of differentially localized proteolytic fragments.

2.2 Introduction

The MEN1 (Multiple Endocrine Neoplasia Type 1) gene, which encodes the protein menin, is a tumor suppressor that has been well conserved across evolution. However, the absence of functional domains, structural motifs, or significant regions of homology to any known genes or proteins affords little insight into the evolutionary origins of MEN1 and the molecular function of

menin176. Menin is thought to be primarily a nuclear protein, targeted by nuclear localization

signals (NLS) in the carboxyl (C)-terminus177. In the nucleus, menin acts as a molecular scaffold

to bind a range of transcriptional activators and repressors to control gene expression in response

to numerous cell signaling pathways111,178. In addition to roles as transcriptional regulators, many

tumor suppressors are known to act as cytoplasmic molecular scaffolds to integrate cell-cell

signaling events179, including synaptogenesis26. Nuclear exit signals (NES) in the amino (N)-

terminus have also been found to shuttle menin into the cytoplasm180, although its role here remains

poorly understood.

Our group was the first to identify a synaptogenic role for MEN1 with a well-conserved orthologue

from the invertebrate mollusk Lymnaea stagnalis. We have previously reported that neurotrophic

factor (NTF)-mediated expression of Lymnaea-MEN1 (L-MEN1/L-menin, described in subsequent text as MEN1/menin) is necessary and sufficient for in vitro and in situ synapse formation between

Lymnaea neurons99,102. Other studies from murine models have demonstrated that menin induces

plasticity in the spinal cord dorsal horn to produce neuropathic pain in response to peripheral nerve

injury105-107, indicating that the role for MEN1 in CNS synapse formation and plasticity has been

conserved across evolution. Despite prevalent MEN1 expression in both the developing and adult

central nervous system (CNS)181, the molecular function of menin in neurons remains, however, unidentified.

In the present study, I took advantage of a MEN1-dependent excitatory cholinergic synapse between Lymnaea CNS neurons to pursue identification of the molecular mechanisms underlying the synaptogenic effect of menin. Here, I report that menin is cleaved at an evolutionarily conserved calpain site in response to NTF signaling. The resulting menin proteolytic fragments coordinate nuclear and synaptic events necessary for excitatory synaptogenesis, including subunit- selective nicotinic acetylcholine receptor (nAChR) gene induction via the N-terminal fragment,

and postsynaptic clustering of nAChR via the C-terminal fragment. This study (i) is the first to demonstrate that menin is cleaved by calpain, (ii) characterizes a role for menin in the transcriptional regulation of neuronal nAChRs, (iii) identifies a novel cytoplasmic function for menin in mediating cell-cell interactions, and (iv) identifies a novel synaptogenic mechanism in which a single gene product coordinates the nuclear transcription and postsynaptic targeting of neurotransmitter receptors.

2.3 Methods

2.3.1 Animals and neuronal cell culture

Lymnaea stagnalis were raised under standard conditions in freshwater aquaria at room temperature (RT; ~22°C), and fed a diet of lettuce. Animals (6-8 weeks old for cell culture; >3 months old for protein samples and CM) were sacrificed by removal of the CNS after anesthesia was induced with a 10% Listerine solution (10m). Identified neurons visceral dorsal 4 (VD4) and left pedal dorsal 1 (LPeD1) were isolated from trypsinized CNS by suction applied through a glass pipette and plated on poly-L-lysine coated glass culture dishes, as previously described in detail144,151. Isolated neurons were maintained overnight (~15-20h) in defined media (DM; trophic factor-deficient media; L-15; Life technologies; special order) or CNS conditioned DM (CM; trophic factor-rich media).

2.3.2 Molecular biology

We have recently described the cloning of Lymnaea-MEN1102. c-Myc (Myc; 5’) and hemagglutinin

(HA; 3’) epitope tags were introduced onto this construct using Myc and HA-encoding primers with the KAPA HiFi HotStart ReadyMix PCR kit (Kapa Biosystems). For N-MEN1 (menin M1-

A413) the stop codon TAG was introduced after A1239. For C-MEN1 (menin Q414-V759) the start codon ATG was introduced ahead of C1240. nAChR C lacking the terminal stop codon was cloned from Lymnaea CNS cDNA, generated as previously described102. eGFP was cloned from the pWPI lentiviral construct (Addgene) and inserted in frame downstream of nAChR C. mCherry was

cloned from the pSicoR-Ef1a-mCh lentiviral vector (Addgene) and extended onto the C-MEN1

sequence by site-overlap-extension PCR, which eliminated the C-MEN1 stop codon. Primers are

shown in Table 2.1. Constructs were inserted into pBlueScript SK- (Clontech). Synthetic mRNA was made using mMESSAGE mMACHINE T7 Ultra Transcription kit (Ambion). Kits were used according to manufacturers’ instructions. mRNA was microinjected into LPeD1 neurons with a low resistance glass electrode, as described previously102. Molecular-grade water was microinjected as a vehicle control. Sequence alignments were generated with Clustal Omega

(EMBL-EBI).

Quantitative (q)PCR expression profiling of LPeD1 neurons was performed as we have recently described in detail102. Gene specific primers are shown in Tables 2.2-2.3. Efficiency values for

qPCR primers ranged between 90-110% (R2 = 0.96-0.99). Negative controls and validations were

102 as previously described . Changes in gene expression (relative to LPeD1 + H2O DM, normalized

to 18s rRNA and β-tubulin), and statistical significance was determined using REST-2009182.

2.3.3 Immunocytochemistry and microscopy

Neurons were fixed for 30m with 4% paraformaldehyde and 0.2% picric acid (Sigma-Aldrich) in

1xPBS, and permeabilized for 1h with incubation media (IM) containing 0.2% Triton, 5% fetal

calf serum, and 0.25% fish gelatin in 1xTBS. Primary antibodies (α-menin [Bethyl Laboratories,

A300-105A]; α-5-HT [AbCam, ab16007]; α-FMRFamide183; α-hemagglutinnin [Covance, MMS-

101P], α-c-Myc [Sigma-Aldrich, C3956]) were used at 1:500 in IM for 1h. Secondary antibodies

(Alexa Fluor 488, 546 or 633 [Invitrogen]) were used at 1:100 in IM for 1h. Three 15m washes in

1xPBS were performed after each incubation, and all incubations were performed at RT.

Presynaptic terminals were labeled with 10 μM FM 1-43 (Molecular Probes) by 10x5s bursts induced in VD4 by current injection through a sharp electrode, then washed on ice with cold

1xPBS and fixed as above. Neurons were mounted with MOWIOL containing DAPI. For live cell imaging, LPeD1 neurons were microinjected with C-MEN1-mCherry and nAChR C-eGFP mRNA, and VD4-LPeD1 axon pairs were maintained in DM for 8-24h. Images were acquired at

10m intervals for ≥8h following the addition of CM. Confocal Z-stacks (1 µm) images were acquired using an A1R MP microscope (Nikon) under a CFI Plan Fluor 20x/0.75 MI objective

(Nikon), with motorized positioning systems (Prior Scientific). Fluorophores were excited with

402, 488, 561 and 651 laser wavelengths in series and emissions collected through 450/50, 525/50,

595/50 and 700/75 filter cubes. Imaging parameters were kept the same amongst relevant samples.

Images were acquired with NIS Elements v4.13.00 software (Nikon) and processed with ImageJ

(NIH).

2.3.4 Preparation of protein samples and Western blotting

Lymnaea CNS tissue for protein samples was acutely dissected (25 CNS) or maintained in organ culture (3 CNS, 3 mL DM at RT for 72h). Tissue was immediately frozen on dry ice and stored at

-80°C. CNS tissue was homogenized in lysis buffer containing 1% Triton, 50 mM Tris-HCl, 150 mM NaCl (pH 8.0) and a cOmplete Mini EDTA-free protease inhibitor cocktail tablet (Roche).

Homogenate was incubated on a shaker for 2h at 4°C, centrifuged at 1,400xg for 20m at 4°C and

stored at -80°C. The protocol for subcellular fractionation was adapted from previously described

methods184,185, with minor modifications. 150 Lymnaea CNS were homogenized by 20 strokes

with a loose pestle followed by 10 strokes with a tight pestle in a Dounce homogenizer in

homogenization buffer (HB; 0.3 M sucrose, 25 mM Tris-HCl [pH 7.4]). All buffers contained cOmplete Mini EDTA-free protease inhibitor cocktail (Roche) or Halt protease inhibitor cocktail

(Thermo Scientific), and all steps were performed on ice or at 4°C. Homogenate was centrifuged

at 1,000xg for 5m to pellet crude nuclei (S1, P1). The pellet (P1) was resuspended in HB and

centrifuged again at 1,000xg for 5m (S2, P2). Nuclear Fraction: the crude nuclei pellet (P2) was

briefly homogenized by 10 strokes with a tight pestle in a Dounce homogenizer, then twice washed

with N-buffer (10 mM NaCl, 3 mM MgCl2, 10 mM Tris-HCl [pH 7.4], 0.1% Nonidet-P40, 0.05%

sodium deoxycholate, 10 mM N-ethyl malemide and 0.5% β-mercaptoethanol), and centrifuged at

1,000xg for 5m. The pellet was centrifuged through a 1.8 M sucrose cushion in 10 mM Tris-HCl

(pH 7.4) at 30,000xg for 30m, then washed again with N-buffer and centrifuged at 1,000xg for

5m. The pellet was resuspended in RIPA buffer (150 mM NaCl, 10 mM Tris-HCl [pH 7.4], 1%

Nonidet-P40, 1% sodium deoxycholate, 0.1% SDS, and 10 mM N-ethyl malemide), sonicated briefly, and centrifuged at 12,000xg for 10m to pellet debris. Cytoplasmic Fraction: supernatants

(S1, S2) were combined and centrifuged at 17,000xg for 30m to pellet crude membranes (P3). The

resulting supernatant was the soluble cytoplasmic fraction. Synaptic Fraction: The crude

membrane pellet (P3) was resuspended in 1.2 M sucrose in 10 mM Tris-HCl (pH 7.4) and allowed

to stand on ice for 30m, then separated on a discontinuous sucrose density gradient established

with 1.2, 0.8 and 0.3 M sucrose in 10 mM Tris-HCl (pH 7.4). This was ultracentrifuged for 30m at 100,000xg using a SW41 Ti rotor (Beckman). The interface between 1.2 and 0.8 M sucrose was collected and pelleted at 150,000xg for 1h using a TLS-55 rotor (Beckman). The pellet was resuspended in HB. Fractions were aliquoted and stored at -80°C. Protein extracts (40 µg/lane) were resolved on SDS-PAGE gels and transferred to PVDF membranes (BioRad). Membrane

blocking and antibody incubations (α-menin at 1:2000; α-β-actin [AbCam, ab8227] at 1:20000; α-

β-tubulin [Sigma-Aldrich, T0198] at 1:2000; α-complexin [Synaptic Systems, 122 102] at 1:2000;

α-histone H3 [Millipore, 06-599] at 1:2000) were performed with 5% skim milk powder + 0.1%

Tween-20 in 1xPBS, overnight at 4°C or for 1h at RT. Bands were detected with IRDye-800CW conjugated α-rabbit or α-mouse IgG (Li-Cor Biosciences) at 1:5000 for 1h at RT. Three 15m washes in 1xPBS + 0.1% Tween-20 at RT were performed after each antibody incubation.

Membranes were visualized using a Li-Cor Odyssey infra-red imager and bands were quantified using Odyssey v3.0 software.

2.3.5 Electrophysiology

Intracellular current clamp recordings were used to characterize single cell physiology or in vitro synapses, as previously described in detail101. For single cell recordings ACh was dissolved into

DM (1 µM), and applied using pressure application through a microelectrode (tip opening ~ 1-5

µm; 250 ms pulse, 10 PSI). An excitatory synapse was confirmed by depolarization of LPeD1 at -

60 mV. The mean amplitude of 5 consecutive excitatory postsynaptic potentials (EPSPs) at -100

mV was used as a measure of synaptic efficacy.

2.3.6 Chemicals

Acetylcholine (ACh), phorbol 12-myristate 13-acetate (PMA) and chelerythrine chloride (Ch Cl-)

were purchased from Sigma-Aldrich. PD150606 was purchased from Tocris. KN-92 and KN-93

were purchased from EMD Millipore. U0126 was purchased from Promega. Drugs were dissolved

into DMSO, and 0.1% DMSO vehicle controls were performed for all experiments.

2.3.7 Experimental design and statistical analysis

To ensure reliability and replicability of the results, all reported data are derived from ≥2 independent experiments, conducted with tissue preparations from independent dissection sessions or cells from independent culture sessions. Sample sizes were limited by uncontrollable factors inherent to the cell culturing techniques used in this study. Data analysis, including fluorescence intensity measurements, subcellular distribution incidence, and electrophysiology, was performed blinded by acquisition file number.

Statistical analyses were performed using SPSS Statistics v22. Data distribution was assessed with

Shapiro-Wilk test, and all data sets were normally distributed (P>0.05), with the exception of two

(P<0.05), DM + N-MEN1 + C-MEN1 (Fig. 2.5C) and CM + 20 μM PD150606 (Fig. 2.6B). The non-normal distribution in these two data sets results from a bimodal distribution, reflecting preparations in which synapse formation either was or was not effectively induced/inhibited.

Differences between two data sets were analyzed with Student’s independent samples t-test (2- sided), and differences amongst three or more data sets were analyzed with univariate ANOVA, with Tukey’s HSD post hoc test if variances were equal (Levene’s statistic P>0.05) or Games-

Howell post hoc test if variances were unequal (Levene’s statistic P<0.05). Incidence data were assessed with Pearson’s Chi-squared test (2-sided). Differences in relative gene expression were determined via pair wise fixed reallocation randomization test using REST-2009.

2.4 Results

2.4.1 Menin is localized to a synapse between Lymnaea neurons

The in vitro reconstructed synapse between the Lymnaea presynaptic interneuron VD4 and its

postsynaptic target LPeD1 is a well-characterized model for the study of both NTF- and MEN1-

dependent synaptogenesis. VD4-LPeD1 forms an inhibitory cholinergic synapse in the absence of

NTF (cultured in defined media; DM), and an excitatory cholinergic synapse in the presence of

NTF (cultured in CNS-conditioned DM; CM, NTF-rich; upregulates MEN1 mRNA expression102).

While menin is expressed in both pre- and postsynaptic neurons, only postsynaptic expression is

required for synaptogenesis. Excitatory cholinergic synapse formation between VD4-LPeD1 in

CM is inhibited by MEN1 knockdown in LPeD199, and can also be induced in DM by MEN1

expression in LPeD1102. Our previous observations therefore suggest that the synaptogenic

function of MEN1 may be specific to the regulation of neurotransmitter receptors, although how

this occurs remains unknown.

12 nAChR subunits have been identified in Lymnaea (L-nAChR A-L, described in subsequent text

as nAChR A-L), which form excitatory or inhibitory nAChRs (cation- or anion-selective channels)158. Given that menin is described as a nuclear protein in the cancer literature177, my

initial hypothesis was that menin induces the transcriptional upregulation of excitatory nAChR

subunits to mediate excitatory synaptogenesis in response to NTF. Therefore, I first sought to

characterize the subcellular localization of menin in neurons. VD4-LPeD1 were cultured in an

axon-axon configuration in CM, and immunocytochemistry (ICC) was performed using a

commercial antibody against a well conserved menin C-terminal epitope. Contrary to my expectations, few neurons showed primarily nuclear localization of menin (Fig. 2.1A; n=5/28,

18%). The majority of pairs exhibited a presumptive synaptic localization (Fig. 2.1B; n=19/28,

68%), whereas the remainder showed primarily cytoplasmic, non-localized distribution (Fig. 2.1C; n=4/28, 14%). I next sought to verify that the axonal clustering of menin depicts synaptic recruitment. To this end, I first labeled FMRFamide neuropeptides expressed in VD4, but not

LPeD1, to visualize presynaptic innervation of the LPeD1 axon. Menin immunoreactivity (see Fig.

2.1B) mimics the axonal distribution patterns of FMRFamide immunoreactivity (Fig. 2.1D; n=4), suggesting that menin clustering occurs at innervated sites along the LPeD1 axon. Secondly, to confirm that menin clustering occurs at functional synaptic sites, presynaptic active zones in VD4 were loaded with the membrane impermeant dye FM 1-43 by depolarization-induced synaptic vesicle recycling, and ICC was then used to label serotonin (5-HT; the neurotransmitter used by

LPeD1) at the postsynaptic membrane, as well as menin (Fig. 2.1E; n=4). Multiple sites of co- localization were observed, indicating that menin clusters to functional synaptic sites where the presynaptic active zone (FM 1-43 positive) is opposed by the postsynaptic membrane (5-HT positive). Menin positive, but FM 1-43 and 5-HT negative sites in VD4 indicate that presynaptic menin does not cluster to synaptic sites. Menin puncta in LPeD1, however, were all FM 1-43 and

5-HT positive, and it is therefore likely that the synaptic menin signal represents postsynaptic recruitment. These data suggest that there is a previously unidentified function for menin in the postsynaptic density.

Figure 2.1 - Postsynaptic clustering of menin in Lymnaea neurons

Figure 2.1: Postsynaptic clustering of menin in Lymnaea neurons

(A-C). ICC characterization of VD4-LPeD1 axon-axon pairs cultured in CM. (A). Nuclear localization of menin (n=5/28). The location of the nucleus was confirmed with DAPI. Arrows, nuclei. (B). Synaptic localization of menin (n=19/28). Arrows, synaptic sites. (C). Non-localized menin (n=4/28). (D). ICC localization of FMRFamide (VD4, presynaptic), shown to illustrate presynaptic innervation (n=4). Arrows, presynaptic sites. (E). Co-localization of FM 1-43 (VD4, presynaptic) and serotonin (5-HT; LPeD1, postsynaptic) to verify that menin clustering occurs at synaptic sites (n=4). Menin puncta at synaptic sites (FM 1-43 and 5-HT positive) are observed along both the LPeD1 axon (arrows) and the VD4 axon (asterisk), due to reciprocal outgrowth of synaptic processes. Menin puncta at non-synaptic sites (FM 1-43 and 5-HT negative) are observed in VD4 (arrowhead), but not LPeD1, suggesting that the source of the synaptic menin signal is postsynaptic. Scale bars, 50 µm.

2.4.2 Menin is cleaved by calpain and the resulting fragments are differentially localized

within neurons

To verify the specificity of menin antibody binding in Lymnaea preparations, I performed ICC

fluorescence intensity analysis on LPeD1 neurons and Western blot (WB) analysis of protein samples from Lymnaea CNS. Firstly, LPeD1 neurons were cultured in DM, CM, or DM + synthetic

MEN1 mRNA microinjection. Relative to DM, ICC fluorescence was increased in LPeD1 neurons

cultured in CM or DM + MEN1 mRNA microinjection (Fig. 2.2A; n=4 each; P=0.001, 0.011

respectively, one-way ANOVA; Table 2.4), indicating an upregulation of menin protein

expression and specificity of the α-menin signal observed in cultured Lymnaea neurons. Secondly,

an appropriate molecular weight band was apparent for Lymnaea menin (84.5 kDa) in WBs, but I

was also struck by the consistent presence of a more rapidly migrating band of ~40 kDa (Fig.

2.2Bi; n=6, representative blot). As a broad-spectrum protease inhibitor was present during sample

preparations, I concluded that this lower band is likely a menin proteolytic fragment generated

endogenously. I next used subcellular fractionation and WB analysis to determine the subcellular

distribution of menin and the presumptive C-terminal proteolytic fragment (C-menin), reasoning

that proteolytic cleavage of the epitope region might explain both the predominant absence of a

nuclear α-menin signal and the predominant synaptic α-menin signal observed with ICC (see Fig.

2.1). CNS microsomes exhibited menin localized to the cytoplasmic fraction and the C-menin

fragment to the synaptic fraction (Fig. 2.2Bii; n=4, representative blot), suggesting that the

synaptic localization I observed (see Fig. 2.1B) represents selective recruitment of the C-menin

fragment.

In mammalian preparations a rapidly migrating band of ~20 kDa is routinely reported with the

menin C-terminal epitope antibody, however this lower band is often dismissed as being due to a

non-specific cross-reaction177. A BLAST search of the C-terminal Lymnaea and mammalian

menin sequences detected by the C-terminal epitope antibody revealed that no other proteins bear

sufficient sequence identity to indicate potential cross-reactions (data not shown). By contrast, my

observations suggest that these faster migrating bands represent specific antibody binding to C-

menin proteolytic fragments present in both invertebrates and vertebrates, likely generated at an

evolutionarily conserved protease consensus site. Upstream of the Lymnaea menin sequence

expansion accounting for the size difference of the lower bands I identified a region of interest

(ROI) of 24 well conserved residues (Fig. 2.3), corresponding to an exposed unstructured loop178, which would be an accessible site for proteolytic cleavage (Fig. 2.2Ci). I evaluated the ROI for conserved protease consensus sequences that would produce menin cleavage fragments in accordance with the observed banding patterns, and identified a putative cleavage site for the calcium-dependent protease calpain within the ROI (Fig. 2.2Cii) using a calpain substrate prediction algorithm186. This site is conserved in menin orthologues from Drosophila to human

(Fig. 2.4), suggesting a strong evolutionary pressure for the maintenance of menin proteolytic

fragments. Calpain cleavage at this site would produce C-menin fragments with predicted

molecular weights of 38 kDa (Lymnaea), and 19 kDa (mammalian), which are in agreement with

the observed banding patterns. I next used CNS organ culture with a cell-permeable calpain inhibitor (20 µM PD150606) and WB analysis to determine whether calpain inhibition would shift band distribution towards full-length menin. Consistent with the predicted calpain cleavage site, I observed an increase in full-length menin and a corresponding decrease in the C-menin fragment

upon calpain inhibition (Fig. 2.2D; n=8 each; C-menin/menin relative fluorescence; DM, 1.17 ±

0.04; DM + 20 µM PD150606, 1.05 ± 0.03; P=0.023, independent t-test).

Calpain-mediated menin cleavage is particularly intriguing considering our previous finding that

LPeD1 neurons exhibit spontaneous activity-dependent calcium oscillations in response to NTFs,

which are required for the functional expression of excitatory nAChRs101. I next generated a MEN1

construct with 5’ (N-terminal) Myc and 3’ (C-terminal) HA epitope tags (Fig. 2.5A), reasoning

that calpain-dependent menin cleavage in response to NTF-induced calcium oscillations would

result in separation of the epitope tags. Synthetic mRNA was microinjected into LPeD1 neurons

cultured in DM, CM, or CM + 20 μM PD150606, and the subcellular distribution of Myc and HA

epitopes was determined with ICC (Fig. 2.5B-C; n≥7; Table 2.5). In CM I observed separation of

the Myc and HA signals, with the Myc epitope localized predominantly in the nucleus, and the

HA eptiope localized predominantly in the cytoplasm. This separation reveals that the

corresponding N-terminal menin fragment (N-menin) is maintained within neurons following

proteolytic cleavage, and is specifically targeted to the nucleus. Calpain inhibition decreased

nuclear Myc eptiope localization and increased cytoplasmic colocalization of Myc and HA

epitopes, indicative of a shift towards full-length menin (P=0.003, one-way ANOVA). In DM I

observed predominantly cytoplasmic overlapping Myc and HA signals, indicative of full-length menin (P<0.001). H2O vehicle control microinjected neurons showed a weak background signal

indicative of non-specific cell surface labeling, where Myc and HA signals were mostly

overlapping, membrane-limited, and distinct from the distribution patterns observed with mRNA

positive samples. Thus, it appears that NTF signaling, in addition to activating the molecular

signals for MEN1 gene induction102, simultaneously provides the high concentration of

intracellular calcium101 required for calpain activation and menin proteolytic cleavage. When taken

together, my observations suggest that the differential targeting of menin, N-menin and C-menin within neurons may serve to mediate distinct functions during synaptogenesis.

Figure 2.2 - Menin is cleaved by calpain

Figure 2.2: Menin is cleaved by calpain

(A). ICC of menin in LPeD1 cultured in DM (i; n=4), CM (ii; n=4), or DM + MEN1 mRNA (iii; n=4). Scale bar, 20 µm. Summary data, fold-increase in menin fluorescence, relative to DM (iv).

Asterisk, statistical significance (one-way ANOVA), P<0.05-0.001. (B). WB of menin (top) in a

Lymnaea CNS whole protein sample (i; n=6, representative blot) and subcellular fractions (ii; n=4,

representative blot). N denotes nuclear fraction, C denotes cytoplasmic fraction, S denotes synaptic

fraction. β-Tubulin (βTUB), complexin (CPLX), and histone H3 (HH3) are shown to verify the

subcellular fractions (bottom). (C). 3D crystal structure of human menin (accession no. 3U84)

showing the conserved ROI (yellow, arrow), an exposed unstructured loop (i; see also Fig. 2.3).

Predicted calpain cleavage site in the conserved ROI of Lymnaea menin (ii; see also Fig. 2.4). Blue

line indicates threshold for significant prediction scores, blue bar and asterisk indicate a significant

predicted calpain cleavage site. (D). WB of menin in protein samples from Lymnaea CNS

incubated in control conditions (DM + 0.1% DMSO) or in the presence of a cell-permeable calpain

inhibitor (DM + 20 μM PD150606) (i; n=8 each, representative blot). β-actin (βACT) is shown to

verify equal loading. Summary data, calpain inhibition reduces menin cleavage (ii). Asterisk,

statistical significance (independent t-test), P<0.05. Error bars, SEM.

Figure 2.3 - The menin conserved region of interest 64

Figure 2.3: The menin conserved region of interest

Alignment of Lymnaea (accession no. AF395538) and mouse menin (accession no.

NM001168488), generated with Clustal Omega (EMBL-EBI). The C-terminal epitope recognized by the menin antibody used in this study is underlined in green. The conserved ROI is shaded green. The stretch of 24 highly conserved residues occurs upstream of the Lymnaea menin sequence expansion that accounts for the size difference observed for Lymnaea and mammalian menin fragments (WB lower bands, see Fig. 2.2). Locations of known functional sequences of menin (NLS, light blue bars; NES, dark blue bars; leucine zipper motif, shaded red) are also shown.

Figure 2.4 - Predicted calpain cleavage site in the menin region of interest is conserved

Figure 2.4: Predicted calpain cleavage site in the menin region of interest is conserved

(A). Sequence alignment of the menin conserved ROI (shaded green) from vertebrate species human, mouse, and zebrafish, and invertebrate species Lymnaea and Drosophila. (B). Multiple

Kernel Learning prediction identifies a presumptive calpain cleavage site in the conserved ROI of all sequences (arrow in A). Blue line indicates threshold for significant prediction scores, blue bars indicate predicted calpain cleavage sites. Asterisks indicate statistical significance of the conserved site, P<0.001.

Figure 2.5 - Menin fragments are differentially targeted within LPeD1 neurons

(A). Schematic of the Myc and HA epitope tagged MEN1 mRNA construct (top) and the predicted

menin protein and fragments (bottom). (B). ICC of Myc and HA localization in Myc-MEN1-HA

mRNA microinjected LPeD1 neurons cultured in CM (i; n=10), CM + 20 μM PD150606 (ii; n=7),

DM (iii; n=10), or vehicle control in DM (iv; n=15; H2O microinjected). Scale bars, 20 µm. (C).

Summary data, relative distribution of the Myc epitope signal in LPeD1. Asterisk, statistical significance (one-way ANOVA), P<0.01-0.001. Error bars, SEM.

2.4.3 Postsynaptic recruitment of the C-menin fragment requires neurotrophic factor- and

activity-dependent signaling

We have characterized a variety of receptor tyrosine kinase (RTK)- and activity-dependent mechanisms required for excitatory cholinergic synaptogenesis between VD4-LPeD1101,102,187,188.

If excitatory synaptogenesis were dependent upon postsynaptic clustering of the C-menin fragment, I reasoned that synaptic recruitment would exhibit kinase-dependent regulation in alignment with some of these previously identified mechanisms. To this end, VD4-LPeD1 axon- axon pairs were cultured in the presence or absence of various kinase inhibitors and menin ICC was used to screen for candidate molecular pathways underlying the synaptic recruitment of C- menin (n≥12 each). In VD4-LPeD1 cultured in CM, menin was mostly localized to the synapse

(Fig. 2.6Ai; Table 2.6, CM data is repeated here from Fig. 2.1), whereas DM showed primarily non-localized menin (Fig. 2.6Aii; P=0.003, Chi-squared test, synaptic distribution relative to CM reported in text). With calpain inhibition, neurons showed reduced synaptic and increased cytoplasmic menin (Fig. 2.6Aiii; CM + 20 µM PD150606; P<0.001), which is consistent with my observations of menin and C-menin subcellular localizations (see Figs. 2.2-2.5), and suggests that the synaptic α-menin signal specifically depicts the C-menin fragment. With disruption of the mitogen-activated protein kinase (MAPK) cascade downstream of RTK signaling, menin was primarily non-localized (Fig. 2.6Aiv; CM + 40 µM U0126; P=0.003). With inhibition of the activity-dependent kinase calcium/calmodulin-dependent protein kinase II (CaMKII), menin fails to localize to the synapse (Fig. 2.6Av; CM + 1 µM KN-93; P=0.013), but exhibits distribution patterns near identical to CM with the inactive analogue (Fig. 2.6Avi; CM + 1 µM KN-92;

P=0.828). Neither a protein kinase C (PKC) inhibitor in CM (Fig. 2.6Avii; CM + 1 µM Ch Cl-;

P=0.348) nor a PKC activator in DM (Fig. 2.6Aviii; DM + 100 nM PMA; P=0.004) affected the synaptic recruitment of menin. These observations indicate that phosphorylation of the C-menin fragment is likely required for synaptic targeting, and that this phenomenon is regulated by the same NTF-induced signaling mechanisms required for excitatory synaptogenesis between VD4-

LPeD1 (references above).

Figure 2.6 - Postsynaptic recruitment of C-menin requires neurotrophic factor- and

activity-dependent signaling

Figure 2.6: Postsynaptic recruitment of C-menin requires neurotrophic factor- and activity-

dependent signaling

(A). ICC of menin in axon-axon paired VD4-LPeD1 cultured in CM (i; NTF-rich; n=28), DM (ii;

NTF-poor; n=12), CM + 20 µM PD150606 (iii; calpain inhibitor; n=18), CM + 40 µM U0126 (iv;

MAPK cascade inhibitor; n=12), CM + 1 µM KN-93 (v; CaMKII inhibitor; n=12), CM + 1 µM

KN-92 (vi; inactive analogue; n=17), CM + 1 µM Ch Cl- (vii; PKC inhibitor; n=15), or DM + 100 nM PMA (viii; PKC activator; n=17). Arrows, synaptic sites. Scale bars, 50 µm. (B). Summary data, subcellular distribution of menin. Asterisks, statistical significance between distribution frequencies, relative to CM (Chi-squared test), P<0.05-0.001.

2.4.4 The C-menin fragment mediates postsynaptic consolidation of excitatory nAChR

The evolutionary conservation of the menin ROI and calpain cleavage site suggests that there is a specific function for one or both of the fragments that is essential to the molecular actions of menin.

To dissect the functional roles of N- and C-menin fragments in MEN1-dependent synaptogenesis,

I generated 5’-truncated (C-MEN1) and 3’-truncated (N-MEN1) constructs at the end of the ROI

adjacent to the calpain cleavage site (Fig. 2.7A). MEN1 mRNA produces menin, N-MEN1 mRNA

produces the N-menin fragment, and C-MEN1 mRNA produces the C-menin fragment. To characterize changes in synaptic physiology induced by the menin fragments, VD4-LPeD1 were paired in a soma-soma configuration in DM or CM, and LPeD1 was microinjected with synthetic

MEN1, N-MEN1, C-MEN1, or N-MEN1 + C-MEN1 mRNA (Fig. 2.8). I performed intracellular recordings to determine the incidence of excitatory synaptogenesis and measured the amplitudes of EPSPs as an indication of synaptic strength (Fig. 2.7A-C; n≥8; Table 2.7). No excitatory synapses were observed in DM + H2O vehicle control (P<0.001, Chi-squared test, relative to CM).

LPeD1 mRNA microinjection of C-MEN1 or MEN1 in DM induced full rescue of the incidence

of excitatory synaptogenesis (P>0.05). mRNA microinjection of N-MEN1 or N-MEN1 + C-MEN1

in DM induced partial rescue of excitatory synaptogenesis, as the incidence of excitatory synapses

was significant relative to DM (P=0.006, 0.002), but also CM (P=0.002, 0.010). N-MEN1 + C-

MEN1 or MEN1 mRNA microinjection in DM induced full rescue of EPSP amplitudes (P>0.05,

one-way ANOVA), whereas EPSP amplitudes were reduced in N-MEN1 or C-MEN1 mRNA

microinjected samples (P=0.001, <0.001). These observations suggest, on the one hand, that there

may be some redundancy in the function of menin, N-menin and C-menin in excitatory

synaptogenesis, but on the other, that the independent molecular functions of the C-menin and N- menin fragments alone are insufficient for the full induction of excitatory synaptogenesis.

The observation that the C-menin fragment induced full synaptogenic rescue but only ~50% EPSP amplitudes supports a requirement for NTF-dependent C-menin synaptic recruitment in postsynaptic development (see Fig. 2.6). I therefore hypothesized that overexpression of the C-

menin fragment, but not the N-menin fragment, in LPeD1 would potentiate excitatory synaptic

transmission in the presence of NTF. The incidence of excitatory synapse formation between VD4-

LPeD1 was unaffected by mRNA microinjection in CM (Fig. 2.7B; P>0.05, Chi-squared test).

Consistent with the above hypothesis, C-MEN1, N-MEN1 + C-MEN1, or MEN1 (which provides the C-menin ‘precursor protein’) mRNA microinjection induced ~2-fold synaptic potentiation

(Fig. 2.7C; P=0.044, 0.031, 0.033, one-way ANOVA), but N-MEN1 mRNA microinjection did not (P=1.000). These data support a role for C-menin synaptic recruitment in the postsynaptic consolidation of excitatory nAChRs.

I next sought to determine whether excitatory synaptogenesis is dependent upon the generation of menin proteolytic fragments. While CM induces bursting activity in LPeD1, DM-cultured LPeD1 neurons also exhibit spontaneous activity188, which could be sufficient for low levels of menin

proteolytic cleavage and underlie MEN1-induced excitatory synaptogenesis in the absence of

NTFs. To this end, intracellular recordings were made from VD4-LPeD1 pairs cultured in CM,

CM + 20 µM PD150606, or DM + MEN1 mRNA + 20 μM PD150606 (Fig. 2.9A-B; n≥12; Table

2.8). Relative to CM, calpain inhibition in CM reduced excitatory synapse formation (P=0.030,

Chi-squared test) and also inhibited MEN1-induced excitatory synaptogenesis in DM (P<0.001).

Synaptic transmission was also reduced upon calpain inhibition in both CM and DM + MEN1

mRNA (P<0.001, one-way ANOVA). Recordings were next made from single LPeD1 neurons to

determine whether the functional expression of excitatory nAChRs in the postsynaptic neuron

induced by MEN1102 is also contingent upon the generation of proteolytic fragments (Fig. 2.9C-D;

n≥10; Table 2.8). Single LPeD1 neurons cultured in DM exhibited an inhibitory response to

exogenous ACh application, indicative of the functional expression of anionic nAChRs. Single

LPeD1 neurons cultured in CM exhibited an excitatory response to ACh (P<0.001, Chi-squared

test, relative to DM), indicative of the functional expression of cationic nAChR. DM + MEN1

mRNA induced the expression of cationic nAChRs in LPeD1 (P=0.007). This nAChR anionic to

cationic functional switch was not dependent upon the production of menin fragments, as an

excitatory response to ACh was observed in LPeD1 cultured in CM + 20 μM PD150606 (P<0.001), as well as DM + MEN1 mRNA + 20 μM PD150606 (P=0.002). Taken together, these data suggest that, while full length menin can induce the functional expression of excitatory nAChRs, MEN1-

dependent excitatory synaptogenesis requires generation of the C-menin fragment to mediate the

postsynaptic recruitment of excitatory nAChRs.

Figure 2.7 - The C-menin fragment mediates postsynaptic consolidation of excitatory nAChRs

Figure 2.7: The C-menin fragment mediates postsynaptic consolidation of excitatory nAChRs

(A). Schematic of MEN1, N-MEN1, and C-MEN1 mRNA constructs (i). Soma-soma paired VD4-

LPeD1 (ii). Scale bar, 20 µm. Inhibitory synapse in DM, action potentials induced in VD4 elicit

hyperpolarization (-50 mV holding potential) in LPeD1 (iii). Excitatory synapse in CM, VD4

activity elicits depolarization (-60 mV holding potential) in LPeD1 (iv). MEN1 mRNA

microinjection in LPeD1 induces an excitatory synapse in DM (v). Representative EPSP traces (-

100 mV holding potential) illustrate MEN1–induced (DM + MEN1) and MEN1-potentiated (CM

+ MEN1) EPSPs (vi). (B). Summary data, incidence of VD4-LPeD1 excitatory synapse formation

in DM + H2O (n=14), DM + N-MEN1 (n=17), DM + C-MEN1 (n=13), DM + N-MEN1 + C-MEN1

(n=20), DM + MEN1 (n=15), CM + H2O (n=43), CM + N-MEN1 (n=13), CM + C-MEN1 (n=12),

CM + N-MEN1 + C-MEN1 (n=8), and CM + MEN1 (n=9). mRNA was microinjected only into

LPeD1 (see also Fig. 2.8). ND, excitatory synapses not detected. Asterisks, statistical significance

relative to CM (Chi-squared test), P<0.05-0.001. ǂ, statistical significance relative to DM, P<0.05-

0.001. (C). Summary data, mean EPSP amplitudes of VD4-LPeD1 synapses, as in (B). ND, EPSPs

not detected. Error bars, SEM. Asterisks, statistical significance (one-way ANOVA), P<0.05-

0.001.

Figure 2.8 - Synthetic mRNA-induced menin expression is specific to microinjected neurons

Axon-axon paired VD4-LPeD1 neurons were cultured in CM, and LPeD1 was microinjected with

HA-tagged C-MEN1 mRNA (n=3). ICC was used to detect endogenous menin (α-menin) and the

C-menin fragment (α-HA). HA epitope tagged C-menin is detected only in the LPeD1 neuron.

Note that nuclear localization of the C-menin fragment is observed coincident with a ‘non- localized' distribution pattern of endogenous menin, where synaptic recruitment has not been promoted by CM. Scale bar, 50 μm.

Figure 2.9 - The C-menin fragment is required for postsynaptic consolidation but not functional expression of excitatory

nAChRs 79

Figure 2.9: The C-menin fragment is required for postsynaptic consolidation but not functional

expression of excitatory nAChRs

(A). Incidence of VD4-LPeD1 excitatory synapse formation in CM + 0.1% DMSO (n=12), CM +

20 µM PD150606 (n=20), or DM + MEN1 + 20 µM PD150606 (n=14; mRNA was injected only

into LPeD1). Asterisk, statistical significance relative to CM (Chi-squared test), P<0.05-0.001.

(B). Mean EPSP amplitudes, as in (A). Error bars, SEM. Insert shows representative EPSP traces

(-100 mV holding potential). Asterisk, statistical significance (one-way ANOVA), P<0.001. (C).

Incidence of excitatory nAChR expression in single LPeD1 cultured in DM + H2O (n=13), CM +

0.1% DMSO (n=12), CM + 20 µM PD150606 (n=13), DM + MEN1 (n=10), or DM + MEN1 + 20

µM PD150606 (n=14). Asterisk, statistical significance relative to DM (Chi-squared test), P<0.01-

0.001. (D). Representative traces, as in (C). ACh application (1 µM, arrow) in single LPeD1 above

firing threshold (-50 mV holding potential, inhibitory nAChRs, top) or below firing threshold (-60

mV holding potential, excitatory nAChRs, bottom).

2.4.5 Menin fragments coordinate subunit-specific transcriptional upregulation and synaptic

targeting of excitatory nAChR

Considering my evidence for the nuclear localization of menin fragments (e.g. see Figs. 2.1A,

2.5Bi, 2.8), I revisited the initial hypothesis that menin induces the transcriptional upregulation of

excitatory nAChR subunits to prime postsynaptic neurons for excitatory synaptogenesis. To

characterize the transcriptional influence of menin versus the proteolytic fragments on the

expression of excitatory nAChRs, the cytoplasm of LPeD1 neurons, cultured in DM and

microinjected with MEN1, N-MEN1 or C-MEN1 mRNA, was isolated for single-cell qPCR (Fig.

2.10A-C). I also evaluated the expression of nAChRs in LPeD1 cultured in CM to determine the

profile of nAChR expression induced by NTF signaling. LPeD1 neurons cultured in DM + H2O

served as a baseline to which the expression of nAChRs in experimental samples was compared

(n=2-3 independent experiments each; n represents a pooled sample of 12-15 single cells and 3

qPCR triplicate replicates). I tested for all 12 Lymnaea nAChR subunits, and detected consistent

qPCR signals for excitatory subunits C, D, E, G, J and inhibitory subunits B, I, K in LPeD1.

nAChR F, H, L qPCR signals were observed infrequently, and nAChR A was not observed (data

not shown). Relative to DM, MEN1 mRNA microinjection induced the transcriptional

upregulation of excitatory nAChR C and J subunits (Fig. 2.10C; P<0.05-0.001, pair wise fixed

reallocation randomization test; Table 2.9). N-MEN1 upregulated nAChR C and J subunits, and also induced the transcriptional upregulation of endogenous MEN1 expression. C-MEN1 upregulated nAChR C and J without inducing endogenous MEN1 expression. CM induced the transcriptional upregulation of endogenous MEN1 as well as excitatory nAChR C and J subunits, suggesting that NTF-induced transcription of excitatory nAChR subunits likely occurs via menin.

CM and N-MEN1 induced comparatively higher levels of nAChR C subunit expression, whereas

MEN1 and C-MEN1 induced comparatively higher levels of nAChR J. This trend suggests that

under appropriate physiological conditions (i.e. NTF stimulation and calpain activation), the N-

menin fragment (i) amplifies MEN1 expression through feed-forward autoregulation, and (ii)

induces a specific expressional profile of cationic nAChR subunits to promote excitatory

synaptogenesis. To determine whether generation of menin fragments is required for nAChR C

expression, I also cultured LPeD1 neurons in CM + 20 μM PD150606 or DM + MEN1 mRNA +

20 μM PD150606. In line with the above hypothesis, full length menin resulting from calpain inhibition in CM or DM was insufficient to induce transcriptional upregulation of nAChR C.

In light of the C-menin-induced synaptic potentiation that I observed (see Fig. 2.7C), I next sought

to determine whether the excitatory nAChR C subunit upregulated by N-menin was subsequently

targeted to synaptic sites with C-menin. To this end, I generated mCherry-tagged C-MEN1 and

eGFP-tagged nAChR C constructs, and synthetic mRNA was microinjected into LPeD1. Axon-

axon paired VD4-LPeD1 neurons were maintained in DM for 8-24h after mRNA microinjection,

and live cell imaging was performed for ≥8h following the addition of CM (n=8). I observed

nuclear localization of the C-menin fragment in 1/8 (13%) LPeD1 neurons, and the addition of

CM induced nuclear export (Fig. 2.10D). Along the LPeD1 axon, I observed both stable and motile

puncta of C-menin co-localized with nAChR C in 5/8 (63%) VD4-LPeD1 pairs (Fig. 2.10E). These

data parallel the menin synaptic and nuclear distribution patterns observed with ICC (see Fig. 2.4).

Taken together, my data support a model in which (i) NTF-induced activity promotes calpain

activation and proteolytic cleavage of menin; (ii) nuclear targeting of the N-menin fragment further

amplifies MEN1 expression and mediates transcriptional upregulation of the nAChR C subunit, which may be required for the synaptic targeting of excitatory nAChRs; and (iii) NTF-induced

phosphorylation promotes postsynaptic targeting of the C-menin fragment, which is required for

the clustering of excitatory nAChRs (Fig. 2.11).

Figure 2.10 - Menin fragments coordinate subunit-specific transcriptional upregulation

and synaptic targeting of excitatory nAChRs

Figure 2.10: Menin fragments coordinate subunit-specific transcriptional upregulation and

synaptic targeting of excitatory nAChRs

(A). LPeD1 cytoplasm isolation for single-cell qPCR. Scale bar, 20 µm. (B). Schematic of mRNA

constructs. (C). Relative gene expression in LPeD1 cultured in DM + MEN1 (n=2), DM + N-

MEN1 (n=2), DM + C-MEN1 (n=2), CM + 0.1% DMSO (n=3), CM + 20 μM PD150606 (n=3), or

DM + MEN1 + 20 μM PD150606 (n=2). Fold change gene expression is shown relative to expression levels in LPeD1 cultured in DM + H2O (n=3). Each n represents a pooled sample of

12-15 single cells and 3 qPCR triplicate replicates. Dark blue bars depict excitatory nAChR C and

J subunits upregulated by menin/fragments, light blue bars depict other excitatory nAChR

subunits, red bars depict inhibitory nAChR subunits. ND, qPCR signal not detected. Error bars,

SEM. Asterisks, statistical significance (pair wise fixed reallocation randomization test) for MEN1,

nAChR C, and nAChR J, P<0.05-0.001 (shown for clarity; other statistically significant

differences were observed, see Table 2.9). (D-E). Live cell imaging of LPeD1 + C-MEN1-mCherry

+ nAChRC-eGFP, axon-axon paired with VD4, following the addition of CM (n=8). Scale bars,

20 µm. (D). C-menin nuclear export at CM t=0m (i), 30m (ii), 60m (iii) and 90m (iv). (E). Stable

(arrowhead) and motile (arrow) C-menin and nAChR C puncta in the LPeD1 axon at CM t=8.5h

(i), 9h (ii), 9.5h (iii) and 10h (iv).

Figure 2.11 - A model for the coordination of nuclear transcription and postsynaptic

clustering of excitatory nAChRs by menin proteolytic fragments

2.5 Discussion

While previous studies on the role of menin in synapse formation and plasticity have characterized

its functional significance, these have so far been unable to delineate the underlying molecular

mechanisms. My report provides the first evidence for a cytoplasmic function of menin, which

was previously thought to be primarily a nuclear protein, in the postsynaptic targeting of excitatory

nAChRs. This study is also the first to demonstrate that menin influences subunit-specific transcription of nAChRs in neurons. Here, I propose a novel model for excitatory synaptogenesis in which the nuclear transcription and postsynaptic targeting of neurotransmitter receptors is coordinated via differential localization and distinct molecular functions of two proteolytic fragments of a single gene product.

2.5.1 Proteolytic cleavage

NTF-mediated molecular signaling events influence nearly all aspects of the development and function of neuronal circuits, from neurogenesis and proliferation to neuronal outgrowth, synapse formation, maturation and plasticity71. NTFs are well known to induce distinct patterns of neuronal

activity necessary for synapse formation82. In Lymnaea LPeD1 neurons, we have previously shown that CM induces stereotypical activity patterns and calcium oscillations required for the conversion from inhibitory to excitatory nAChRs101,188. Calpains, which are ubiquitously expressed, are one of many classes of effector proteins underlying the signal transduction of calcium as a second

messenger, and calpain-dependent proteolysis has been implicated in a diverse array of calcium-

dependent cellular processes, including apoptosis, adhesion, cytoskeletal reorganization, synaptic

plasticity, and neurodegeneration189. A number of tumor suppressor proteins have been found to

be cleaved by calpain, including p53190, NF2191, PTEN192, as well as the Dlg homolog PSD-95193.

In the present study, I have shown that proteolytic cleavage of menin in neurons allows NTF-

dependent molecular signaling to coordinate nuclear and cytoplasmic events prerequisite for

excitatory synaptogenesis. My observations also support a general model in which the production

of proteolytic fragments may play a common role in the regulation of cell-cell interactions by tumor suppressors.

2.5.2 Transcriptional regulation

Synapse formation and long-lasting forms of synaptic plasticity, involving the growth of new synaptic connections, require activity-dependent gene induction, de novo protein synthesis, and site-specific targeting of effector proteins194. The cyclic AMP response element binding protein

(CREB) cascade is the prototypical mechanism through which extracellular stimuli are transduced

into gene expression changes necessary for synaptic development and remodeling95,122,195-197. The transcription factor CREB is a convergence point for multiple second messenger systems, allowing neurotransmitter-receptor interactions, calcium influx, and impulse activity to cooperatively influence activity-dependent gene transcription98,198. While there is currently no evidence for the

transcriptional activation of MEN1 by CREB, multiple cis-regulatory regions have been identified in the promoter region of human MEN1, and the activity of these regulatory elements is dependent

upon cellular context (endocrine vs. non-endocrine cell type)199, indicating that this facilitates cell

type-specific transcriptional regulation. This could conceivably be extended to explain the

responsiveness of MEN1 gene expression to context-dependent molecular signals in neurons, such

as NTF-activated second messenger systems. For instance, calcium influx through L-type voltage-

gated calcium channels (VGCCs) is critically involved in the activation of immediate early genes,

typically transcription factors, that mediate subsequent transcriptional changes underlying synapse

formation and plasticity200. L-type VGCC signal transduction to the nucleus is mediated via locally

aggregated CaMKII201, which has also been show to activate CREB98. We have previously

reported that inhibition of L-type VGCCs prevents the expression of excitatory nAChRs in

Lymnaea LPeD1 neurons101, and that this can be bypassed by the injection of MEN1 mRNA102.

Considering the findings of the present study, these observations suggest that menin is the molecular intermediary between NTF-induced activity-dependent signaling and the transcriptional upregulation of excitatory nAChR subunits relevant to excitatory synaptogenesis.

Menin binds to both transcriptional activators (e.g. MLL1, SMADs) and repressors (e.g. JunD,

NFκβ) to regulate gene transcription in response to numerous cell signaling cascades111. Menin is also known to bind directly to DNA and influence gene transcription202. Furthermore, the

transcription of MEN1 is influenced by the intracellular levels of menin, indicating a promoter

system that facilitates auto-regulation199. This provides support for my observation that the N-

menin fragment induced endogenous MEN1 expression. The transcriptional influence of menin

and the C-terminal fragment, without nuclear localization, could occur indirectly through the

regulation of other transcription factors. For instance, the nuclear export of menin has been

reported to reduce nuclear accumulation of β-catenin and thereby influence its transcriptional

activity180. Considering that full length menin was localized primarily to the cytoplasm, and that

addition of NTF induced the export of nuclear localized C-menin, my observations suggest that the transcriptional influence of menin in Lymnaea neurons depends primarily upon the nuclear

localized N-menin fragment.

Neuronal nAChRs are diverse and heterogeneous due to the variety of possible subunit

combinations that can comprise the pentameric channels, and the factors that regulate specific

patterns of expression, transcriptional regulation, assembly and trafficking are not well

understood203. Here, I identified MEN1-induced transcriptional upregulation of excitatory nAChR

C and J subunits in Lymnaea LPeD1 neurons, where nAChR C subunit expression was specifically promoted by the N-menin fragment. As Lymnaea nAChR C (α-type subunit) and J (β-type subunit) are not known to form functional homopentamers159, these are likely to be accessory subunits that modulate the assembly, function, or trafficking of nAChRs. I demonstrate here that calpain

inhibition both prevented nAChR C upregulation and excitatory synapse formation between VD4-

LPeD1, but not nAChR J upregulation or excitatory nAChR expression in single LPeD1 neurons.

Taken together, these findings suggest that NTF stimulation induces the transcriptional

upregulation of the nAChR C subunit via N-menin, and raise the possibility that nAChR C

expression may be required for the formation of nAChR channels that are competent for synaptic

targeting.

2.5.3 Postsynaptic recruitment

In this study, I describe for the first time the functional significance of the C-menin proteolytic fragment and its NTF-dependent postsynaptic targeting. Phosphorylation of human menin at C-

terminal residues Ser543 and Ser583 has been previously identified, and this was found to have no

effect on nuclear localization or transcriptional regulation204. The phosphorylation of these C-

terminal serine residues and the absence of a transcriptional effect provides support for my

observation that the postsynaptic recruitment of C-menin requires phosphorylation by the Ser/Thr kinases MAPK or CaMKII. As the corresponding serine residues are conserved in Lymnaea menin,

I suspect that this may serve as the molecular signal for synaptic localization of the C-menin fragment.

The synaptic potentiation induced by C-menin and its co-localization with nAChR is analogous to the effects described for the postsynaptic clustering of glutamate receptors by the molecular scaffold PSD-95 in mammalian neurons27. Menin is also known to act as a molecular scaffold,

mediating crosstalk between multiple signaling pathways in the regulation of gene transcription111.

If this scaffolding function is maintained in the C-menin fragment, my data would raise the

intriguing possibility that the C-menin fragment is the previously unidentified molecular scaffold

for nAChR clustering in neurons. Cholinergic synaptic development at the vertebrate

neuromuscular junction (NMJ) is dependent upon agrin signaling through muscle-specific kinase

(MuSK) receptors to mediate the clustering of nAChRs via the intracellular effector rapsyn31.

While agrin, MuSK, and rapsyn are expressed in neurons and have been found to influence the function of cholinergic synapses in the CNS205,206, rapsyn is not essential for postsynaptic nAChR

clustering at interneuronal synapses33. By contrast, we have previously demonstrated that

postsynaptic MEN1 knockdown inhibits NTF-dependent excitatory cholinergic synaptogenesis99, and that MEN1 induces excitatory cholinergic postsynaptic development in the absence of NTF signaling102. When taken together, our observations illustrate both the necessity and sufficiency

for MEN1 in excitatory cholinergic synaptogenesis between Lymnaea central neurons. In murine

models, independent reports have shown that peripheral nerve injury upregulates both MEN1105-

107 and the modulatory nAChR α5 subunit8 in the spinal cord dorsal horn to produce neuropathic

pain. While the induction of neuropathic pain by menin has been characterized in terms of

glutamatergic hyperexcitability107, this could also be consistent with MEN1-induced cholinergic plasticity resulting in glutamatergic facilitation207, as cholinergic innervations are present in the

dorsal horn208. If the molecular actions of menin in neurons have indeed been conserved across evolution, this may be a MEN1-dependent transcriptional upregulation and synaptic clustering of

α5-containing nAChRs, in response to enhanced NTF signaling following injury209.

2.5.4 Conclusion

Taken together, my observations on the role of calpain-dependent menin cleavage in excitatory

synapse formation elucidate the molecular actions of menin in neurons, and also reveal a novel

synaptogenic mechanism in which a single gene product coordinates the nuclear transcription and

postsynaptic targeting of neurotransmitter receptors. Unraveling the distinct molecular functions

of the menin proteolytic fragments and whether these are tissue specific or context dependent may

ultimately provide new insights into menin’s actions as a tumor suppressor, and further

fundamental knowledge regarding the development of cholinergic synapses in the CNS, as well as

neurodegenerative conditions such as Alzheimer’s disease in which NTF210, nAChR211, and tumor

suppressor function109,110 are compromised.

Table 2.1 - Lymnaea cloning primers

Construct Target Primer Sequence Myc- L-menin with N-terminal 5’GATGATCTCGAGATGGAGCAGAAGCTGAT MEN1- Myc and C-terminal HA CTCAGAGGAGGACCTG HA epitope tags GCGGGCTTTCGAGACCGAG 3’GATGATGCGGCCGCCTAAGCGTAATCTGG AACATCGTATGGGTAGACTATTTCTCTCCTT GGCC N-MEN1 L-menin C-terminal 5’GATGATCTCGAGATGGAGCAGAAGCTGAT truncation CTCAGAGGAGGACCTG (residues 1-413) GCGGGCTTTCGAGACCGAG 3’GATGATGCGGCCGCCTATGCCCATCCAAC ATGGAGC C-MEN1 L-menin N-terminal 5’GATGATCTCGAGATGCAGCACCTAACATT truncation CTCTCTC (residues 414-759) 3’GATGATGCGGCCGCCTAAGCGTAATCTGG AACATCGTATGGGTAGACTATTTCTCTCCTT GGCC C-MEN1- C-menin 5’GATGATCTCGAGATGCAGCACCTAACATT mCherry (residues 414-759) CTCTCTC with C-terminal 3’CCATGTTATCCTCCTCGCCCTTGCTCACCA mCherry tag TCCCAGACCCAGACCCAGACCCGACTATTTC TCTCCTTGGCCGTTTTCGTGC mCherry C-menin 5’GCACGAAAACGGCCAAGGAGAGAAATAG (residues 414-759) TCGGGTCTGGGTCTGGGTCTGGGATGGTGAG with C-terminal CAAGGGCGAGGAGGATAACATGG mCherry tag 3’GATGATGGATCCTTACTTGTACAGCTCGTC CATGC nAChR L-nAChR C 5’GATGATCTCGAGATGGATGTGCTGACCAG C-eGFP with C-terminal eGFP tag CGC 3’GATGATGAATTCCCCAGACCCAGACCCGT ATATACATGTGTTATTTGGATC eGFP L-nAChR C 5’GATGAATTCATGGTGAGCAAGGGCGAGG with C-terminal eGFP tag 3’GATGGATCCCTAGCTACTAGCTAGTCGAG

Table 2.2 - Lymnaea RT-PCR gene specific primers

Target Accession 5’ Primer Sequence 3’ Primer Sequence Number β Tubulin X15542 TCCTACTTTGTGGAATGG ATGACGAGAATTATGTCA ATCC TTAGAC 18s rRNA Z73984 CTGGTTGATCCTGCCAGT CTTCCGCAGGTTCACCTA AG C L-MEN1 AF395538 TCGAGACCGAGCGAAGA TTTCGTGCAGATCCTGTT AAC GG L-nAChR A DQ167344 GTGTATGCTCGTCGGCAT TCATCCTCGTCCTCCGACT GT T L-nAChR B DQ167345 GGCCTTGACCTGCACTTA CATTCGCGGGCTAGGTAC CC TC L-nAChR C DQ167346 CCAGCGCCATTTTCTTCT TGCCACAGAAGCAAGCTG TC TT L-nAChR D DQ167347 CCTCACGGACAATGGCA GGTGTTTCCGGTTTCGTC GTA AT L-nAChR E DQ167348 TAGTGCCAAGCGGTTGT TATCTGGGCGGATGTTGA ACG GA L-nAChR F DQ167349 CTGCTGATGTCCGTGGTT ACGACCCTCCACTCGTGA GT AT L-nAChR G DQ167350 GGCTCACCATGGAACAA GCAACAAACTGCCACTCT CAA GC L-nAChR H DQ167351 CAGGCTGTCTGGCGTAC ACGCCACTCATTCAGCAC AAC AT L-nAChR I DQ167352 GTGTGCTTCCTGCTTGTG CCCTAATGTTCGTGGCCT GT TC L-nAChR J DQ167354 AGGTTGGGATGGCAGAA CCCGATGAGAGTGACCAA GTG CA L-nAChR K DQ167353 GGATTCATTGATGGCTTC TGACGCACAGTGAGGTGA AACA TG L-nAChR L DQ167355 CCAGATTCCGCTGGCTCT CTCATCGAATGGCAACCT TA CA

Table 2.3 - Lymnaea qPCR gene specific primers

Target Accession 5’ Primer Sequence 3’ Primer Sequence Efficiency Number (%) β Tubulin X15542 ATCCAGGAGCTCTT CTGTGAACTCCATC 105.80 CAAGCG TCGTCC 18s rRNA Z73984 CACGGGGAGGTAGT GCCCTCCAATGGGT 103.40 GACG CCTC L-MEN1 AF395538 TGGAGTTCGCTGTC CAAAGGCAACACCA 93.52 TCGAAG AAGCAA L-nAChR A DQ167344 CGGCCGATACTCAA GCCAGATGTTGGTG 103.88 CGAGTC TGGATG L-nAChR B DQ167345 GCCAATGTCTGCAG GTCGCTTTGTTCCTG 100.42 CAGAC CACGG L-nAChR C DQ167346 GAGACGGACATGAT CATAGGTCCTGCCG 109.04 CAAGCC ACGGC L-nAChR D DQ167347 GGCCTCACAGGACT GGTCAGAGGCGTTG 106.21 ACCAAC TACACG L-nAChR E DQ167348 GAGGAGGAGTGGCT CATGATTTGGTTCTT 102.55 ACAAC CTCGTC L-nAChR F DQ167349 5’GGCCTGTCACTCA GTCGTACTGTTCTAT 90.88 TTCAGATC ATCCCAC L-nAChR G DQ167350 GACCAAGTCTTGGT GATAGGAGGAGCCA 108.01 TTCTGG ATGAGG L-nAChR H DQ167352 CGCTCTACGTTTCCA GAAGGAGACGTGGA 110.91 TCGAG CAGTG L-nAChR I DQ167352 GTACCGCTTCCAGT GTTGACCACTGGCC 93.52 GATATC TGATG L-nAChR J DQ167354 GGAAGGACTACCAG CATCGGCATTGTTG 106.21 CTGGAG AAAAGCAC L-nAChR K DQ167353 CTTCCGGCGTAGGT GCAGGTGATTCTGG 96.06 CCACC AGGTATC L-nAChR L DQ167351 GGCAGACCCGAACG CACAAGTGCCAGGC 102.17 AATACC CAAAGG

Table 2.4 - Immunocytochemistry fluorescence of menin in LPeD1

Treatment Mean P 1 Relative n Fluorescence Fluorescence 2 ± SEM (AU) ± SEM DM 418.98 ± 15.84 (F=14.465; P=0.002) 1.00 ± 0.05 4 CM 674.88 ± 36.93 0.001 1.61 ± 0.07 4 DM + MEN1 mRNA 606.28 ± 44.99 0.011 1.45 ± 0.08 4

1. One-way ANOVA with Tukey’s HSD post hoc test, significance (P) relative to DM

2. Fluorescence values (AU; arbitrary units) relative to DM

Table 2.5 - Immunocytochemistry fluorescence of c-Myc-tagged menin in LPeD1

Treatment Relative Fluorescence 1 ± SEM P 2 n

CM 1.25 ± 0.11 (F=24.610; P<0.001) 10 CM + 20 µM PD150606 0.67 ± 0.09 0.003 7 DM 0.49 ± 0.03 <0.001 10

1. Fluorescence values (AU) of nuclear c-Myc signals normalized to cytoplasmic c-Myc

signals

2. One-way ANOVA with Tukey’s HSD post hoc test, significance (P) relative to CM

Table 2.6 - Subcellular distribution of menin in VD4-LPeD1 axon-axon pairs

Treatment Incidence (n) χ2 P 1 Nuclear CM + 0.1% DMSO 5/28 - - DM 0/12 2.449 0.118 CM + 20 µM PD150606 3/18 0.011 0.917 CM + 40 µM U0126 0/12 2.449 0.118 CM + 1 µM KN-93 1/12 0.598 0.440 CM + 1 µM KN-92 2/17 0.299 0.585 CM + 1 µM Ch Cl- 0/15 3.031 0.082 DM + 100 nM PMA 1/17 1.313 0.252 Non-Localized CM + 0.1% DMSO 4/28 - - DM 10/12 17.603 <0.001 CM + 20 µM PD150606 14/18 18.544 <0.001 CM + 40 µM U0126 10/12 17.603 <0.001 CM + 1 µM KN-93 8/12 10.975 0.001 CM + 1 µM KN-92 4/17 0.618 0.432 CM + 1 µM Ch Cl- 7/15 5.380 0.020 DM + 100 nM PMA 12/17 14.634 <0.001 Synaptic CM + 0.1% DMSO 19/28 - - DM 2/12 8.827 0.003 CM + 20 µM PD150606 1/18 17.305 <0.001 CM + 40 µM U0126 2/12 8.827 0.003 CM + 1 µM KN-93 3/12 6.234 0.013 CM + 1 µM KN-92 11/17 0.047 0.828 CM + 1 µM Ch Cl- 8/15 0.882 0.348 DM + 100 nM PMA 4/17 8.318 0.004

1. Chi-squared test (2-sided), significance (P) relative to CM + 0.1% DMSO

Table 2.7 - Incidence of VD4-LPeD1 excitatory synapse formation and EPSP amplitudes

Treatment Incidence of χ2 P 1 Excitatory Synapses (n) DM + H2O 0/14 29.524 <0.001 2 CM + H2O 35/43 - - DM + N-MEN1 mRNA 7/17 9.384 0.002 2 DM + C-MEN1 mRNA 10/13 0.126 0.722 2 DM + N-MEN1 + C-MEN1 mRNA 10/20 6.593 0.010 2 DM + MEN1 mRNA 10/15 1.387 0.239 2 CM + N-MEN1 mRNA 10/13 0.126 0.722 2 CM + C-MEN1 mRNA 9/12 0.240 0.624 2 CM + N-MEN1 + C-MEN1 mRNA 7/8 0.173 0.677 2 CM + MEN1 mRNA 8/9 0.292 0.589 2 Treatment Mean EPSP P 3 n Amplitude ± SEM (mV) 3 DM + H2O ND - 0 CM + H2O 10.57 ± 0.67 (F=21.64; P<0.001) 35 DM + N-MEN1 mRNA 4.38 ± 0.83 0.001 7 DM + C-MEN1 mRNA 5.50 ± 0.65 <0.001 10 DM + N-MEN1 + C-MEN1 mRNA 7.61 ± 1.03 0.338 10 DM + MEN1 mRNA 7.20 ± 0.86 0.101 10 CM + N-MEN1 mRNA 10.71 ± 0.43 1.000 10 CM + C-MEN1 mRNA 18.51 ± 1.90 0.044 9 CM + N-MEN1 + C-MEN1 mRNA 20.71 ± 2.09 0.031 7 CM + MEN1 mRNA 18.56 ± 1.76 0.033 8

1. Chi-squared test (2-sided), significance (P) relative to CM

2. Differences were also significant relative to DM + H2O (Chi-squared test, 2-sided),

P<0.05-0.001

3. One-way ANOVA with Games-Howell post hoc test, significance (P) relative to CM

4. ND indicates excitatory synapses and EPSPs were not detected

Table 2.8 - EPSP amplitudes of VD4-LPeD1 synapses and incidence of excitatory nAChR expression in single LPeD1 neurons

Treatment (VD4-LPeD1) Incidence of Excitatory χ2 P 1 Synapses (n) CM + 0.1 % DMSO 11/12 - - CM + 20 µM PD150606 11/20 4.693 0.030 DM + MEN1 mRNA + 20 µM 2/14 15.476 <0.001 PD150606 Treatment (VD4-LPeD1) Mean EPSP Amplitude P 2 n ± SEM (mV) CM + 0.1 % DMSO 11.28 ± 0.92 (F=33.42; 11 P<0.001) CM + 20 µM PD150606 2.78 ± 0.65 <0.001 11 DM + MEN1 mRNA + 20 µM 1.91 ± 0.38 <0.001 2 PD150606 Treatment (LPeD1) Incidence of Excitatory χ2 P 3 nAChR Expression (n) DM + H2O 1/13 - - CM + 0.1 % DMSO 11/12 17.629 <0.001 CM + 20 µM PD150606 10/13 12.764 <0.001 4 DM + MEN1 mRNA 6/10 7.304 0.007 4 DM + MEN1 mRNA + 20 µM 9/14 9.258 0.002 4 PD150606

1. Chi-squared test (2-sided), significance (P) relative to CM

2. One-way ANOVA with Tukey’s HSD post hoc test, significance (P) relative to CM

3. Chi-squared test (2-sided), significance (P) relative to DM

4. Differences were also not significant relative to CM + 0.1% DMSO (Chi-squared test, 2-

sided), P>0.05

Table 2.9 - Relative gene expression in LPeD1 neurons

LPeD1 Fold-Change qPCR Target Standard Error P 3 Treatment Expression 1,2 DM + MEN1 1 0.77 - 1.30 - H2O nAChR C 1 0.77 - 1.30 - nAChR D 1 0.52 - 1.95 - n=3 nAChR E 1 0.89 - 1.13 - nAChR G 1 0.90 - 1.12 - nAChR J 1 0.74 - 1.38 - nAChR B 1 0.85 - 1.19 - nAChR I 1 0.58 - 1.80 - nAChR K 1 0.75 - 1.35 - DM + MEN1 13,468.13 10,383.84 - 15,880.83 0.036 MEN1 mRNA nAChR C 77.11 60.19 – 99.76 0.023 nAChR D 0.20 0.11 - 0.34 <0.001 n=2 nAChR E 0.41 0.37 - 0.44 0.030 nAChR G 0.91 0.82 - 1.04 0.367 nAChR J 30,432.67 23,465.23 - 38,361.50 <0.001 nAChR B 0.54 0.40 - 0.73 <0.001 nAChR I 0.64 0.40 - 1.03 0.207 nAChR K 2.90 2.28 - 3.59 0.011 DM + MEN1 3,264.68 2,600.10 - 4,026.46 <0.001 N-MEN1 nAChR C 1,458.23 1,199.52 - 1,815.40 <0.001 mRNA nAChR D 0.21 0.11 - 0.38 0.036 nAChR E 0.15 0.12 - 0.18 0.066 n=2 nAChR G 0.54 0.46 - 0.63 0.093 nAChR J 21.96 16.21 - 26.79 <0.001 nAChR B 0.95 0.46 - 2.39 0.884 nAChR I 0.26 0.15 - 0.47 0.066 nAChR K 0.26 0.20 - 0.35 0.060 DM + MEN1 1.66 1.34 - 2.16 <0.001 C-MEN1 nAChR C 184.22 144.94 - 216.81 <0.001 mRNA nAChR D 0.10 0.06 - 0.18 <0.001 nAChR E 0.40 0.35 - 0.43 0.063 n=2 nAChR G 4.32 4.04 - 4.78 0.033 nAChR J 1,114.10 863.99 - 1,378.61 <0.001 nAChR B 0.02 0.02 - 0.02 0.029 nAChR I ND 4 - - nAChR K 14.65 11.16 - 17.46 <0.001

100

Table 2.9 Continued – Relative gene expression in LPeD1 neurons

LPeD1 Fold-Change qPCR Target Standard Error P 3 Treatment Expression 1,2 CM + MEN1 3,672.08 3,026.69 - 4,661.74 <0.001 0.1% DMSO nAChR C 2,034.64 1,674.87 - 2,468.81 <0.001 nAChR D 2.67 1.38 - 4.77 <0.001 n=3 nAChR E 1.24 1.06 - 1.45 0.134 nAChR G 21.67 18.19 - 26.46 <0.001 nAChR J 163.17 128.41 - 204.93 <0.001 nAChR B 0.72 0.30 - 2.21 0.687 nAChR I 1.01 0.69 - 1.63 0.924 nAChR K 40.37 26.96 - 68.60 <0.001 CM + MEN1 749.72 581.40 - 907.07 <0.001 20 μM nAChR C 3.07 2.48-3.67 0.075 PD150606 nAChR D ND 4 - - nAChR E 0.82 0.72 - 0.95 0.066 n=3 nAChR G 2.27 1.73 - 2.89 0.072 nAChR J 19.27 14.69 - 24.30 0.060 nAChR B 0.26 0.21 - 0.32 0.018 nAChR I 3.55 2.32 - 5.61 0.060 nAChR K 8.15 6.45 - 10.46 <0.001 DM + MEN1 19,222.26 14,796.54 - 24,588.00 <0.001 MEN1 mRNA nAChR C 0.007 0.00 - 0.02 0.020 20 μM nAChR D 0.185 0.10 - 0.30 0.011 PD150606 nAChR E 0.43 0.39 - 0.48 0.041 nAChR G 1.985 1.10 - 2.94 0.095 n=2 nAChR J 24.525 17.59 - 36.44 <0.001 nAChR B 0.016 0.01 - 0.02 <0.001 nAChR I ND 4 - - nAChR K 3.148 2.35 - 4.33 <0.001

1. Expression normalized to 18s rRNA and β-Tubulin reference genes

2. Fold-change expression relative to LPeD1 – DM + H2O

3. Pair wise fixed reallocation randomization test (REST-2009), significance (P) relative to

LPeD1 – DM + H2O

4. ND indicates the transcript was not detected in the qPCR reaction.

101

Chapter Three: Tumor suppressor menin is required for subunit-specific nAChR α5

transcription and nAChR-dependent presynaptic facilitation in cultured mouse

hippocampal neurons

Sections of this chapter have been submitted for publication in the following manuscript:

Getz A, Xu F, Visser F, Persson R & Syed NI (2016). Tumor suppressor menin is required for subunit-specific nAChR α5 transcription and nAChR-dependent presynaptic facilitation in cultured mouse hippocampal neurons.

3.1 Abstract

In the central nervous system (CNS), cholinergic transmission induces synaptic plasticity events required for learning and memory. However, our understanding of these circuits is hindered because the factors regulating specific patterns of neuronal nicotinic acetylcholine receptor

(nAChR) subunit expression and functional clustering remain unidentified. Recent studies from our group have implicated calpain-dependent proteolytic fragments of the synaptogenic factor menin, the product of the MEN1 tumor suppressor gene, in subunit-specific transcriptional regulation and synaptic targeting of neuronal nAChRs in the invertebrate CNS. Here, I sought to determine whether the molecular mechanisms underlying menin's cholinergic synaptogenic functions have been conserved in the vertebrate CNS. My data from primary mouse hippocampal neuron cultures demonstrate that the C-terminal menin fragment co-localizes with α7 subunit- containing nAChRs at glutamatergic presynaptic terminals. MEN1 knockdown induced a specific reduction in nAChR α5 subunit expression, and also eliminated nicotine-induced presynaptic facilitation. Furthermore, I found that MEN1 transcription and menin proteolytic cleavage were

102

disrupted in a mouse model of Alzheimer’s disease (AD). Taken together, my results suggest that the synaptogenic function of menin occurs via the regulation of nAChR subunit composition and functional clustering, and identify menin perturbations as a potential mechanism underlying the cholinergic synaptic dysfunction in AD.

103

3.2 Introduction

Synapse formation and synaptic plasticity in the CNS require the coordination of nuclear

transcription and site-specific targeting of nascent synaptic proteins in response to extracellular

(e.g. neurotrophic factors71) and cell-cell (e.g. neuroligin-neurexin44,47, neurotransmitter-

receptor212) signaling interactions. Mounting evidence from our group99,102,213 and others105-107

supports the evolutionarily conserved function of menin, the protein product of the MEN1

(multiple endocrine neoplasia type 1) tumor suppressor gene176, in mediating specific synapse

formation and synaptic plasticity. The molecular mechanisms underlying the synaptogenic

function of menin, however, have not been well characterized. Menin is a multifunction scaffold

protein that integrates extracellular and cell-cell signaling interactions, intracellular molecular

cascades, and nuclear transcription111. One particularly enigmatic aspect of menin’s function is

that, whereas its expression is ubiquitous, it acts as a tumor suppressor only in certain cell types214.

Considering that MEN1 expression is found in most tissues, it seems that as-yet uncharacterized molecular actions of menin may be deemed necessary for specialized cellular functions that extend beyond its basic biological roles in genome maintenance215 and cell cycle regulation216.

We have recently reported that the menin orthologue from the invertebrate mollusk Lymnaea stagnalis (L-MEN1/L-menin) acts postsynaptically, via the coordinated actions of two

differentially localized proteolytic fragments, to induce the subunit-specific transcriptional

upregulation and synaptic clustering of nAChRs during cholinergic synaptogenesis between

central neurons213. This work gave rise to two important questions regarding menin’s neuronal

104

molecular functions that remained unanswered: (i) in parallel with the conserved synaptogenic

function of menin, are the molecular mechanisms underlying the function of menin in neurons also

conserved across evolution; and (ii) does menin act ubiquitously as a synaptogenic factor, or are

its actions specific to the regulation of neuronal nAChRs?

In the mammalian CNS, the activation of cholinergic synapses induces synaptic modulation and

plasticity events which are critically required for learning and memory217. Disruptions of

cholinergic projections, their synaptic connections, and nAChR expression and function are early

events central to the cognitive decline that occurs in neurodegenerative AD218,219. Despite the

importance of cholinergic synaptic activity to cognition, learning and memory, our understanding

of the mechanisms governing the assembly, function and maintenance of central cholinergic

synapses is incomplete, because defining the role of cholinergic synaptic function and the

regulation of nAChRs in the CNS has been technically challenging. Specifically, dissection of the properties governing cholinergic circuits is complicated by the fact that (i) the projections are diffuse, (ii) excitatory and inhibitory target neurons receive cholinergic innervation on both pre and postsynaptic terminals, (iii) nAChR channels are highly heterogeneous, and (iv) the molecular scaffolds underlying the spatial segregation of distinct nAChR subtypes to presynaptic terminals

or dendritic arbors are unidentified220.

While the molecular components of the signaling cascades (agrin-MuSK-rapsyn and neuregulin-

ErbB) governing cholinergic synaptogenesis at the neuromuscular junction (NMJ)76 are expressed

in central neurons and have been shown to modulate the function of central cholinergic

synapses205,206, rapsyn is not required for the functional clustering of neuronal nAChRs33. It

105

therefore follows that central neurons must employ an as-yet unidentified postsynaptic scaffold to

cluster nAChRs, and menin is a promising candidate molecule. The highest levels of MEN1

expression in the CNS are found in the hippocampus181, a center for learning and memory which

receives extensive cholinergic innervation and exhibits abundant nAChR expression221.

Considering our observations from an invertebrate model that postsynaptic targeting of an L-menin proteolytic fragment mediates the functional clustering of neuronal nAChRs213, there exists an

intriguing possibility that one of the specialized cellular functions of menin is the regulation of

nAChR expression and clustering in neurons.

In the present study, I employed mouse CNS tissue and hippocampal neuron cultures to explore

whether the molecular actions of menin in neurons are tied to the regulation of nAChR expression

and function in the mammalian CNS. Here, I report that menin exhibits calpain-dependent

proteolytic cleavage and differential subcellular distribution of the resulting fragments. The C-

terminal proteolytic fragment co-localized extensively with α7 subunit-containing nAChRs at glutamatergic presynaptic terminals, suggesting that the C-terminal menin fragment acts as a molecular scaffold for the axonal targeting and functional clustering of α7-nAChRs in neurons. I also found that menin exhibits cell type-specific expression, present in neurons, but not glia, in hippocampal cultures. MEN1 knockdown in vitro selectively inhibited the transcription of nAChR

α5, but otherwise induced a general transcriptional upregulation. MEN1 knockdown also eliminated nicotine-induced presynaptic facilitation of glutamatergic transmission, but did not affect baseline glutamatergic synaptic function. This suggests that the synaptogenic actions of

menin are specific to the regulation of nAChR subunit composition and functional clustering.

Furthermore, I found that MEN1 transcription and menin proteolytic cleavage were disrupted in a

106

mouse model of AD. Taken together, my results (i) support an evolutionarily conserved mechanism in which menin regulates the subunit-specific transcriptional regulation and synaptic targeting of nAChRs, (ii) suggest that the synaptogenic actions of menin are specific to the regulation of neuronal nAChRs, and (iii) identify menin perturbations as a potential mechanism contributing to the cholinergic synaptic dysfunction observed in AD.

107

3.3 Methods

3.3.1 Animals and neuronal cell culture

All animal procedures were approved by the University of Calgary institutional animal use and

care committee, in accordance with the standards established by the Canadian Council on Animal

Care. Brains and tissue were dissected from 2.5 month old (mo) male C57/BL6 mice (Charles

River) that were anesthetized with isoflurane and sacrificed by decapitation. Hippocampi were

dissected in ice cold artificial cerebral spinal fluid from 6 mo male or female 5xFAD +/- and -/- mice that were anesthetized with CO2 and sacrificed by decapitation. 5xFAD mice were generously

provided by Dr. Peter Stys. Tissue was immediately frozen on dry ice and stored at -80°C.

Dissociated primary hippocampal neuron cultures were prepared from embryonic day 18 (e18)

C57/BL6 mice (Charles River). Pregnant dams were anesthetized with isoflurane and sacrificed by decapitation. E18 embryos were immediately dissected and sacrificed by decapitation.

Hippocampi were dissected in ice cold 1xHBSS containing 10 mM HEPES (310 mOsm, pH 7.2), and treated with an enzyme mixture containing papain (50 U/mL), 150 mM CaCl2, 100 µM L-

cysteine, and 500 µM EDTA in neurobasal medium (NBM) for 20m at 37°C, then washed 3x with

neurobasal media (NBM) supplemented with 4% FBS, 2% B27, 1% penicillin-streptomycin and

1% L-Glutamine (GIBCO). Neurons were dissociated by trituration with polished glass Pasteur pipettes, and plated at a low density onto glass coverslips (washed with nitric acid and coated with

poly-D-lysine (30 µg/mL; Sigma Aldrich) and laminin (2 µg/mL; Sigma Aldrich)) in costar 12

well plates (VWR) in NBM supplemented as above. The next day the culture media was changed

to NBM supplemented with 2% B27, 1% penicillin-streptomycin and 1% L-Glutamine. Neurons

108

° were maintained at 37 C with 5% CO2, and ~50% of the media was changed every 3-4 days.

Neurons were cultured in control conditions (as above), with 0.1% DMSO vehicle control, or with

20 μM PD150606 (Tocris) dissolved in DMSO.

3.3.2 Molecular biology

Western blotting (WB) and subcellular fractionation was performed as previously described213.

RNA samples were obtained from day in vitro (DIV) 7 cultured mouse hippocampal neurons with

the RNeasy Plus Micro kit, and from 5xFAD hippocampi with the RNeasy Plus Mini kit (Qiagen).

cDNA was synthesized with the QuantiTect Reverse Transcription kit (Qiagen) and purified with

a NucleoSpin PCR Clean-up column (Macherey-Nagel). Quantitative (q)PCR was performed with

SYBRgreen (Qiagen) and primers directed to a region of 80-120 bases. Kits were used according

to manufacturers’ instructions. Intron spanning gene specific primers are shown in Table 3.1.

Efficiency values for qPCR primers ranged between 85-110% (R2 = 0.97-1.00). Negative controls

and validations were as previously described102. Relative gene expression (normalized to β-actin and β-tubulin reference genes), and statistical significance was determined using REST-2009182.

3.3.3 Lentivirus production and transduction of neuronal cultures

Small hairpin (sh)RNA-encoding constructs were designed against MEN1 or a non-target control

(NTC) sequence, and cloned into pLL3.7 (Addgene). MEN1 shRNA sequence:

CCGGTACCACTGTCGCAACCGAAATCTCGAGATTTCGGTTGCGACAGTGGTATTTTG.

NTC shRNA sequence: CGCGATAGCGCTAATAATTTCTCGAGAAATTATTAGCGCTATC

109

GCGCTTTTTG. HEK-293 cells were cultured in DMEM supplemented with 10% FBS and 1%

° penicillin-streptomycin (GIBCO) and maintained at 37 C with 5% CO2. HEK-293 cells were transfected with a mixture of pLL3.7 containing the MEN1- or NTC-shRNA, along with psPAX2 and pMD2.G (Addgene) using HEPES/CaPO4 precipitation. After 4h the media was replaced with

14 mL fresh media. After 24h, lentivirus-containing media was harvested and replaced with 14 mL fresh media. Harvested media was centrifuged at 5,000 rpm for 5m, filtered and stored at 4°C,

for a total of three harvests. The harvested media was ultracentrifuged for 2h at 50,000xg using a

SW 28 Ti rotor (Beckman). Lentivirus pellets were resuspended in 1xPBS and stored in aliquots

at -80°C. Viral titers were determined using serial dilutions in HEK-293 cells. Transduction

efficiency was estimated by visualization of transduction marker green fluorescent protein (GFP)

fluorescence after 24h indicating viral titers of ~109 IU/mL. Mouse hippocampal neurons were transduced with MEN1-shRNA or NTC-shRNA encoding lentivirus after 1 DIV by spinoculation

(2m at 2,000 rpm) using a multiplicity of infection of ~0.2, and the media was changed after 24h.

GFP fluorescence was observed after ~24-48 hours, and transduction efficiency of mouse hippocampal neurons was estimated to be ~60-80%. MEN1 knockdown was confirmed by qPCR and immunocytochemistry (ICC).

3.3.4 Immunocytochemistry and microscopy

Cultured mouse hippocampal neurons were fixed at DIV 3-14 for 30m with 4% paraformaldehyde

and 0.2% picric acid (Sigma Aldrich) in 1xPBS, and permeabilized for 1h with incubation medium

(IM) containing 0.5% Triton and 10% goat serum in 1xPBS. Primary antibodies (α-menin C-

terminal epitope [Bethyl Laboratories; A300-105A]; α-nAChR α5 [Abcam; ab41173]; α-menin N-

110

terminal epitope [Santa Cruz Biotechnology; sc-374371]; α-synaptotagmin [EMD Millipore;

MAB5200]; α-PSD-95 [Antibodies Incorporated; 75-028]; α-GFP [Invitrogen; A11120, A11122];

α-NeuN [EMD Millipore; MAB377]; α-GFAP [Abcam; 16997]) were used at 1:500 in IM for 1h. nAChRs were labeled with Alexa Fluor 555 conjugated α-Bungarotoxin (Invitrogen; B35451) at

2 µg/mL in IM for 1h. Secondary antibodies (Alexa Fluor 488, 546, or 633 conjugated goat α- rabbit or α-mouse [Invitrogen]) were used at 1:100 in IM for 1h. Three 15m washes in 1xPBS were performed after each incubation, and all incubations were performed at room temperature.

Neurons were mounted using ProLong Gold antifade reagent with DAPI (Invitrogen). Images were collected from ≥2 samples prepared from independent culture sessions.

Confocal Z-stack imaging was performed as previously described213. Super resolution imaging was performed with fluorescence super-resolution imaging modality Conical Diffraction

Microscopy (CoDiM), which consists of a beam-shaping unit and a reconstruction processing algorithm222. The CoDiM100 module (BioAxial) was coupled to a C2 confocal microscope under

an Apo TIRF 60x Oil DIC N2 objective (Nikon). Fluorophores were excited with 488, 561 and

640 laser wavelengths and emissions collected through 525/50, 598/44 and 710/50 filter cubes.

The fibered output of the laser source of the microscope was diverted through the CoDiM100

beam-shaper module. The laser beam was thereafter coupled to the input port of the scanning head

of the microscope. Four differently shaped intensity patterns were utilized to illuminate the sample.

Each pattern has a topology which gains access to independent structural information of the

sample. The fluorescent light was collected on an Orca Flash 4 V2 sCMOS camera (Hamamatsu

Photonics). Each scanning point in the sample plane was hence interrogated four times by the

excitation light, generating four different micro images. To obtain the super-resolved image the

111

micro images were then processed by a dedicated algorithm. Image processing and fluorescence

intensity analysis were performed with ImageJ (NIH).

3.3.5 Electrophysiology

Whole-cell patch-clamp recordings of spontaneous synaptic currents were made at a holding

potential of -70 mV in the presence of 0.5 µM tetrodotoxin (TTX) from DIV 10-14 untreated

control or lentivirus transduced mouse hippocampal pyramidal neurons expressing the GFP

transduction marker, as previously described for rat cortical cultures223. Nicotine (Sigma Aldrich)

was dissolved into the external recording solution (10 µM) and applied using pressure application

through a microelectrode (tip opening ~ 1-5 µm; 250 ms pulse, 10 PSI). Nicotine-induced facilitation was determined by analyzing the frequency and amplitude of spontaneous synaptic events that occurred during a 10s interval before and after the nicotine pulse. Data was analyzed using Mini Analysis v6.0.7 (Synaptosoft Inc).

3.3.6 Experimental design and statistical analysis

To ensure reported results were reliable and replicable, data were derived from ≥2 independent experiments using cells from independent culture sessions or tissue preparations from separate animals collected during independent dissection sessions. Data analyses were performed blinded by acquisition file number. Statistical analyses were performed using SPSS Statistics v22 for

Windows. Differences in fluorescence intensity or co-localization for immunocytochemistry (ICC) were assessed with univariate ANOVA with Games-Howell post hoc test (Levene’s statistic

P<0.05) or Student’s independent samples t-test (2-sided). Amplitude and frequency of synaptic

112

events were assessed with univariate ANOVA with Tukey’s HSD post hoc test (Levene’s statistic

P>0.05), and nicotine-induced facilitation was assessed with Student’s paired samples t-tests (2- sided). Significant differences in relative gene expression were determined via pair wise fixed reallocation randomization test using REST-2009182.

113

3.4 Results

3.4.1 Menin proteolytic cleavage fragments are differentially localized within neurons

Menin contains both nuclear localization signals (NLS) and nuclear export signals (NES), which are known to shuttle menin in and out of the nucleus in response to various signal transduction cascades180,224. We have recently characterized the proteolytic cleavage of L-menin at a calpain consensus sequence within a stretch of 24 highly conserved residues that forms an exposed unstructured loop213. The calpain cleavage site and 24 residues are nearly identical in menin orthologues from Drosophila to human213, suggesting that there has been considerable evolutionary selection for menin proteolytic fragments, and that these fragments perform necessary biological functions that are yet to be fully characterized. In Lymnaea CNS neurons, we found that the carboxyl (C)-terminal L-menin fragment localized to the postsynaptic membrane and promoted the functional clustering of nAChRs during cholinergic synaptogenesis213. As a synaptogenic function for menin is found from invertebrates to vertebrates, I hypothesized that the underlying molecular mechanisms, including the generation and differential subcellular localization of proteolytic fragments, would be conserved in mammalian neurons. I first performed

WB analysis and subcellular fractionation on protein samples from mouse brain tissue. Cleavage at the evolutionarily conserved calpain consensus site would produce mouse menin fragments with predicted molecular weights of 48.5 kDa (amino (N)-terminal menin fragment; N-menin) and 19 kDa (C-terminal menin fragment; C-menin). Using commercial menin antibodies against C- terminal and N-terminal epitopes (Fig. 3.1), I detected bands of appropriate molecular weight for menin (67.5 kDa), as well as faster migrating lower bands of ~19 kDa (α-C-terminal menin) and

114

~48.5 kDa (α-N-terminal menin) (Fig. 3.2A; n≥3 each, representative blots). Considering that (i)

a broad-spectrum protease inhibitor was present during sample preparation, (ii) the size of the

lower bands are in agreement with the predicted banding patterns of calpain cleavage fragments,

and (iii) the cumulative size of the lower bands corresponds to the molecular weight of full length

menin, I propose that these faster migrating bands represent endogenous menin proteolytic

fragments resulting from calpain cleavage. I then performed subcellular fractionation and WB

analysis, and found that mouse brain microsomes exhibited full length menin predominantly in the

nuclear fraction, the N-menin fragment in the cytoplasmic fraction, and the C-menin fragment in

the synaptic fraction (Fig. 3.2B-C; n=6, representative blots).

Next, I performed immunocytochemistry (ICC) on primary mouse hippocampal cultures (DIV 7)

to determine whether the patterns of menin subcellular distribution identified by WB are also

observed in situ. ICC with the N- and C-terminal menin antibodies showed both α-N- and α-C-

terminal menin signals present in neuronal nuclei, indicative of full length menin. α-N-terminal signals were restricted to the nuclear and perinuclear compartment, whereas α-C-terminal signals also exhibited punctate staining along neurites (Fig. 3.2D; n=13). This separation of menin epitopes in neurons is consistent with the subcellular fractionation results, and supports my hypothesis that menin is proteolytically cleaved and the fragments are differentially localized in neurons. It should also be noted that I found menin expression in hippocampal cultures to be confined to neurons, as both α-N- and α-C-terminal menin signals were absent in the glial layer

(identified by α-NeuN and α-GFAP ICC) that is prominent in mature cultures (Fig. 3.3; n≥4; DIV

14).

115

To verify that the α-C-terminal menin signal in neurites depicts the calpain-dependent proteolytic fragment, neurons were cultured in the presence of a cell-permeable calpain inhibitor (20 μM

PD50606) or vehicle control (0.1% DMSO) and processed for ICC as above (DIV 7). Consistent with the predicted calpain cleavage site and the observed subcellular localizations, the appearance of C-menin puncta in neurites was reduced upon calpain inhibition (Fig. 3.4 A-B; n≥10). Relative to vehicle control, nuclear N-menin fluorescence was unaffected by calpain inhibition (Fig. 3.4C;

P=0.238, independent t-test; Table 3.2), while neurite C-menin fluorescence was significantly reduced (Fig. 3.4D; P<0.001). Taken together, these data demonstrate that menin is proteolytically cleaved by calpain, and that the C-terminal fragment is subsequently targeted into the neurites of hippocampal neurons.

I also performed WB analysis of protein samples from various mouse tissue samples to determine whether the ~19 kDa C-menin fragment might be present outside of the CNS. I found that the C- menin fragment was similarly abundant in non-neuronal excitable tissues such as skeletal muscle and heart, but was less prevalent in non-excitable tissues (Fig. 3.5; n=3). This observation supports the role of the Ca2+-dependent protease calpain in menin proteolytic cleavage, and suggests that the molecular functions of the C-menin fragment may not be specifically restricted to neurons.

116

Figure 3.1 - N- and C-terminal epitopes recognized by the menin antibodies

Mouse menin sequence (accession no. AF016398). The N-terminal epitope region recognized by the N-menin antibody is underlined in red. The C-terminal epitope region recognized by the C- menin antibody is underlined in green. The conserved ROI is underlined in black, and the calpain cleavage site is indicated with an arrow213. Locations of nuclear localization signals (dark blue, double bar) and nuclear exit signals (dark blue, single bar) are also shown.

117

Figure 3.2 - Menin fragments are differentially localized in CNS neurons

118

Figure 3.2: Menin fragments are differentially localized in CNS neurons

(A). WB of mouse brain protein samples with menin C-terminal (C; left) and N-terminal (N; right) epitope antibodies (n≥3, representative blots), detects full length menin (black arrow), as well as

N-terminal (light grey arrow) and C-terminal (dark grey arrow) menin proteolytic fragments. (B).

WB of subcellular fractions from mouse brain with C-terminal (C; top) and N-terminal (N; bottom) epitope menin antibodies (n=6, representative blot). N denotes nuclear fraction, C denotes

cytoplasmic fraction, S denotes synaptic fraction. Menin localizes to the nucleus, the C-menin

fragment localizes to synaptic membranes, and the N-menin fragment localizes to the cytoplasm.

(C). As in B, histone H3 (HH3), synaptophysin (Syp), and β-tubulin (TUB) are shown to verify

the subcellular fractions. (D). ICC localization of menin in mouse hippocampal neurons at DIV 7

(n=13 images, 4 independent samples), with N-terminal (i) and C-terminal (ii) epitope antibodies,

and the nuclear stain DAPI (iii). (iv) shows merged channels. Arrows illustrate the separation of

α-N-menin and α-C-menin signals. Scale bar, 50 μm.

119

Figure 3.3 - Menin exhibits neuron-specific expression in hippocampal cultures

120

Figure 3.3: Menin exhibits neuron-specific expression in hippocampal cultures

(A). ICC localization of menin in mouse hippocampal neurons at DIV 14 (n=13 images, 4 independent samples), labeled with N-terminal (i) and C-terminal (ii) epitope antibodies, and the nuclear stain DAPI (iii). (iv) shows merged channels. Neurons (arrowhead) were positive, and glia

(arrows) were negative for α-N-menin and α-C-menin signals. (B). ICC localization of neuronal and glial markers in mouse hippocampal neurons at DIV 14 (n=4 images, 2 independent samples), labeled with NeuN (i) and GFAP (ii) antibodies, and the nuclear stain DAPI (iii). (iv) shows merged channels. Neurons (arrowhead) were positive for α-NeuN signals, and glia (arrows) were positive for α-GFAP signals. Scale bars, 20 μm.

121

Figure 3.4 - Menin is cleaved by calpain

122

Figure 3.4: Menin is cleaved by calpain

(A-B). ICC localization of menin in mouse hippocampal neurons at DIV 7, cultured in the presence

of 0.1% DMSO vehicle control (A; n=11 images, 4 independent samples), or 20 μM PD150606, a

cell permeable calpain inhibitor (B; n=10 images, 3 independent samples). Cells were labeled with

menin N-terminal (i) and C-terminal (ii) epitope antibodies, and the nuclear stain DAPI (iii). (iv) shows merged channels. Scale bars, 20 μm. (C-D). Summary data, fluorescence intensity of the nuclear α-N-menin signal was unaffected by calpain inhibition (C), whereas the neurite α-C-menin

signal was reduced (D). ***, statistical significance (independent t-test), P<0.001. Error bars,

SEM.

123

Figure 3.5 - Menin proteolytic fragments are present in non-neuronal tissue

WB of mouse tissue protein samples with the menin C-terminal epitope antibody (n=3, representative blot). Lane 1: Spinal cord; 2: Skeletal muscle; 3: Heart; 4: Lung; 5: Kidney; 6: Liver;

7: Spleen; 8: Pancreas. The ~19 kDa C-menin fragment is prevalent in excitable tissues such as the spinal cord, skeletal muscle and heart (lanes 1-3; dark grey arrow), suggesting that C-menin is generated by Ca2+/calpain-dependent proteolytic cleavage in non-neuronal tissues. The appearance

of slower migrating menin immunoreactive bands (e.g. lane 2, ~80 kDa) suggest that menin is

subject to posttranslational modifications (e.g. palmitoylation?) in a tissue-specific manner, and the appearance of faster migrating menin immunoreactive bands (e.g. lane 2, ~40 kDa; Lane 8,

~50 kDa and ~ 15 kDa) suggest that there are additional, as-yet uncharacterized proteolytic cleavage sites for menin.

124

3.4.2 C-menin co-localizes with α7 subunit-containing nAChRs at glutamatergic presynaptic

terminals

While cholinergic innervation is absent in hippocampal cultures, α-bungarotoxin (α-BTX) sensitive nAChRs, which contain the α7 subunit225, are localized to glutamatergic presynaptic

terminals, and their activation is well known to induce presynaptic facilitation207,226. These reports

suggest that the necessary components of cholinergic postsynaptic machinery underlying

functional expression and clustering are expressed in hippocampal neurons notwithstanding the

absence of cholinergic presynaptic input, which seems to be distinct from the mechanism of

postsynaptic cholinergic development at the neuromuscular junction (NMJ)88,90,227, but consistent

with what we have previously observed in single Lymnaea CNS neurons, where the expression of

nAChRs is dependent upon L-MEN1 expression induced by neurotrophic factor (NTF) signaling

cascades102.

To determine whether the punctate localization of the C-menin fragment along neurites might also indicate nAChR clustering in mammalian hippocampal neurons, I used super resolution fluorescence microscopy on DIV 7 neurons labeled for α-BTX, and antibodies against the C- terminal menin epitope, as well as the synaptic vesicle protein synaptotagmin (Syt) as a presynaptic marker, or the glutamatergic scaffold postsynaptic density 95 (PSD-95), as a postsynaptic marker

(Figs. 3.6-3.7; n≥9). I identified a nearly 1:1 incidence of co-localization between C-menin and α-

BTX labeled puncta. C-menin-positive puncta that were α-BTX-negative, and vice-versa, were rarely observed (<10%). The degree of co-localization of C-menin with nAChRs was significantly higher than the co-localization with either the presynaptic marker Syt or the postsynaptic marker

125

PSD-95 (Fig. 3.6A-C; P=0.006 and P<0.001 respectively, one-way ANOVA; Table 3.3). C-menin

was also more closely associated with presynaptic rather than postsynaptic sites (P=0.007).

Considering our previous observations that the synaptogenic effects of L-MEN1 occur at the postsynapse99,102,213, the low correlation of C-menin and the glutamatergic postsynaptic density

scaffold PSD-95 suggests that menin does not function as a ubiquitous synaptogenic factor, but

that its molecular actions are specific to the clustering of nAChRs. The nearly perfect correlation

of C-menin and α-BTX supports the hypothesis that the C-menin fragment acts as a molecular

scaffold for the clustering of α7-nAChRs, and furthermore, the high degree of correlation with Syt

at presynaptic terminals suggests that the clustering function of C-menin may be specific to the

axonal targeting of α7-nAChR.

126

Figure 3.6 - The C-terminal menin fragment co-localizes with α7 subunit-containing

nAChRs at presynaptic terminals

127

Figure 3.6: The C-terminal menin fragment co-localizes with α7 subunit-containing nAChRs at

presynaptic terminals

(A). Super resolution image of a synaptic ROI at DIV 7 (n=9 images, 2 independent samples),

labeled with α-C-terminal menin (i), α-BTX, to detect α7 subunit-containing nAChRs (ii), and α- synaptotagmin (Syt), to detect presynaptic sites (iii). (iv) shows merged channels. (B). As in A,

only labeled with α-PSD-95 to detect postsynaptic sites (n=10 images, 2 independent samples).

Asterisks illustrate extrasynaptic co-localization of C-menin and nAChRs, arrows illustrate

synaptic co-localization with Syt (A) or PSD-95 (B). Scale bars, 2 μm. (C). Summary data, incidence of co-localization of Syt (presynaptic), PSD-95 (postsynaptic), and α-BTX (nAChR α7) puncta with C-menin. **, statistical significance (one-way ANOVA), P<0.01-0.001. Error bars,

SEM.

128

Figure 3.7 - Field of view images for super-resolution microscopy

129

Figure 3.7: Field of view images for super-resolution microscopy

(A). Mouse hippocampal neurons at DIV 7 labeled with α-BTX to detect α7 subunit-containing nAChRs (i), a C-terminal epitope menin antibody (ii), and a synaptotagmin antibody to detect presynaptic sites (iii). (iv) shows merged channels. Boxed area is the synaptic ROI depicted with the super resolution image in Fig. 3.6A. (B). As in A, only labeled with a PSD-95 antibody to detect postsynaptic sites (iii). Boxed area is the synaptic ROI depicted in the super resolution image in Fig. 3.6B. Scale bars, 50 μm.

130

3.4.3 Menin mediates subunit-selective transcriptional regulation of nAChR α5

Menin is largely regarded as a nuclear protein177, and its dual role as a transcriptional activator and

transcriptional repressor, acting through distinct complexes, is well documented178. We recently

reported that L-menin induces subunit-selective transcriptional upregulation of nAChRs213, and I

hypothesized that menin would similarly mediate nAChR subunit-specific transcriptional

regulation in mammalian neurons. To this end, mouse hippocampal cultures were transduced with

lentivirus constructs encoding non-target control (NTC) or MEN1 shRNA at DIV 1, and RNA samples were collected from untreated control, NTC shRNA- and MEN1 shRNA-transduced samples at DIV 7 for qPCR analysis (Fig. 3.8A; n=6, triplicate replicates; Table 3.4). I assayed for neuronal nAChR subunits α2-7 and β2-4, the glutamate receptor subunits GluR1 (AMPA-type) and NR2A (NMDA-type), as well as the synaptic marker synaptophysin (Syp). NTC shRNA had a minimal effect on transcription, inducing a slight increase in nAChR β2 (Fig. 3.8B; P=0.014, pair wise fixed reallocation randomization test) and a slight decrease in NR2A (P=0.019) transcript abundance. shRNA-mediated knockdown of MEN1 (P=0.009), however, induced a subunit- specific reduction in nAChR α5 transcripts (P=0.002), whereas most other transcripts were elevated (P<0.05-0.001). These data suggest, on the one hand, that nAChR α5 subunit expression requires menin-dependent transcriptional activation, and on the other, that the general increase in transcriptional activity is likely due to the loss of MEN1-mediated transcriptional repression.

Next, I used ICC fluorescence analysis to verify knockdown of menin and downregulation of

nAChR α5 in MEN1 shRNA-transduced cultures at the protein level at DIV 3, 7, 10 and 14 (Fig.

3.9A-F; n≥12 each). Images were obtained from regions containing both lentivirus-transduced

131

(GFP+) and untransduced (GFP-) neurons. For analysis, somal ICC fluorescence values in GFP+

neurons were compared to GFP- neurons. NTC shRNA-transduced neurons exhibited a

GFP+/GFP- fluorescence ratio of approximately 1:1 for C-menin, N-menin, nAChR α5 and α-

BTX at all time points (P>0.05, independent t-test; Table 3.5). In MEN1 shRNA-transduced neurons, C-menin fluorescence in GFP+ neurons was significantly lower than in GFP- neurons at

DIV 3-14 (P<0.01-0001; Table 3.6). Intriguingly, N-menin fluorescence was reduced at DIV 3-7

(P<0.01), but recovered at DIV 10-14 (P>0.05), suggesting that the protein stability of menin, N- and C-menin fragments is regulated independently in response to menin intracellular levels. nAChR α5 fluorescence was reduced at DIV 7-14 (P<0.01-0.001) but not at DIV 3 (P>0.05), suggesting that nAChR α5 downregulation follows MEN1 knockdown with a delay, which likely reflects the half-life of existing α5 subunit transcripts and protein. α-BTX fluorescence was reduced at DIV 3 (P<0.05), but not at DIV 7-14 (P>0.05), despite transcriptional upregulation of nAChR α7 (see Fig. 3.8). These observations suggest that nAChR α5 gene induction requires menin-dependent transcriptional activation, but also that transcriptional perturbations caused by a loss of menin-dependent transcriptional repression are not translated to differences in protein abundance for all affected transcripts.

132

Figure 3.8 - Menin mediates subunit-specific transcriptional regulation of nAChR α5

133

Figure 3.8: Menin mediates subunit-specific transcriptional regulation of nAChR α5

(A). Live cell phase contrast (i-iii) and GFP fluorescence (iv-vi) images of untreated control (i,iv),

NTC shRNA-encoding (ii,v), and MEN1 shRNA-encoding (iii,vi) lentivirus transduced hippocampal cultures at DIV 7 (n=18 images, 6 independent samples). Scale bar, 100 μm. (B).

Summary data, fold change gene expression in mouse hippocampal cultures at DIV 7, relative to untreated control, determined by qPCR (n=6 each, triplicate replicates). MEN1 knockdown reduces nAChR α5 expression. ND, qPCR signal not detected. *, statistical significance (pair wise fixed reallocation randomization test), P<0.05-0.001. Error bars, SEM.

134

Figure 3.9 - Menin knockdown reduces nAChR α5 expression

135

Figure 3.9: Menin knockdown reduces nAChR α5 expression

(A-B). ICC characterization of menin and nAChR expression in NTC shRNA (A) and MEN1 shRNA (B) lentivirus transduced hippocampal cultures at DIV 7 (n≥4 images, ≥2 independent samples). Untransduced neurons were GFP negative (-) and transduced neurons were GFP positive

(+). The expression of menin was determined with C-terminal (i) and N-terminal (ii) epitope antibodies, and the expression of nAChR was determined with a nAChR α5 antibody (iii) and α-

BTX (iv). Left panel shows ICC labels, right panel shows GFP. Scale bars, 50 μm. (C-F).

Summary data, fluorescence intensity in GFP+ neurons relative to GFP- neurons (n≥12). Dashed line represents a 1:1 ratio indicating no change. Asterisks, statistical significance (independent t- test); *, P<0.05. **, P<0.01. ***, P<0.001. Error bars, SEM.

136

3.4.4 Menin is required for the functional expression of nAChRs in hippocampal neurons

The application of nicotine to hippocampal neurons in both slice preparations and in culture is well

known to increase the frequency of spontaneous miniature excitatory postsynaptic currents

(mEPSCs), via the facilitation of neurotransmitter release207. This indicates that Ca2+ influx

through activated nAChRs increases synaptic vesicle release probability at glutamatergic

presynaptic terminals, and corresponds well with our observation that C-menin and α-BTX positive puncta are observed at presynaptic over postsynaptic sites (see Fig. 3.6). Patch-clamp recordings (DIV 10-14) were next made from untreated control, NTC shRNA- and MEN1 shRNA- transduced pyramidal neurons (GFP+; transduced at DIV 1) to evaluate whether knockdown of menin perturbs the functional expression of nAChRs in hippocampal neurons and precludes nicotine-induced presynaptic facilitation (Fig. 3.10-3.12; Table 3.7). NTC shRNA and MEN1 shRNA expression did not alter the baseline amplitude or frequency of mEPSCs, suggesting that glutamatergic synaptogenesis proceeds normally upon menin knockdown (Fig. 3.10A-F; n≥15 each; P=0.844 and P=0.241, one-way ANOVA), despite the transcriptional upregulation of synaptophysin and the glutamate receptor subunits GluR1 and NR2A (see Fig. 3.8). Application of nicotine (10 μM) increased the frequency of mEPSCs in untreated control and NTC shRNA- transduced neurons (P<0.001, paired t-test), however, MEN1 shRNA-transduced neurons did not exhibit nicotine-induced presynaptic facilitation (P=0.628). As α-BTX fluorescence was unchanged (see Fig. 3.9), this observation suggests that the functional clustering of nAChRs at presynaptic terminals is disrupted upon menin knockdown. In all cases, application of nicotine did not affect the amplitude of mEPSCs (P>0.05, paired t-test). Taken together, these data support the hypothesis that menin’s synaptogenic function is specific to the regulation of nAChRs in neurons,

137

and may furthermore be limited to the axonal targeting of α7 subunit-containing nAChRs to presynaptic terminals.

138

Figure 3.10 - MEN1 knockdown eliminates nAChR-dependent presynaptic facilitation

139

Figure 3.10: MEN1 knockdown eliminates nAChR-dependent presynaptic facilitation

(A-B). Representative patch clamp traces from hippocampal pyramidal neurons (DIV 10-14). nAChR-mediated increase in mEPSC frequency (e.g. arrow) is observed in NTC-shRNA expressing neurons (A), but is absent in MEN1-shRNA expressing neurons (B). (C). Summary data, mean frequency of mEPSCs before (Pre-nicotine, 10s) and after (Post-nicotine, 10s) the nicotine pulse (10 μM, 250 ms, 10 PSI), in untreated control (n=19), NTC shRNA (n=15), and

MEN1 shRNA (n=17) transduced neurons. ***, statistical significance (paired t-test), P<0.001

(D). Relative mEPSC frequency. *, statistical significance (one-way ANOVA), P<0.05. (E). Mean amplitude of mEPSCs before and after the nicotine pulse. (F). Relative mEPSC amplitude. Error bars, SEM.

140

Figure 3.11 - Control traces for patch clamp recordings

(A-C). Application of vehicle control (n=6; external recording buffer solution [EBS], 250 ms

pulse, 10 PSI) did not induce facilitation of mEPSC frequency (A; paired t-test, P=0.752) or

amplitude (B; paired t-test, P=0.145). (C). Relative mEPSC frequency and amplitude. (D)

mEPSCs were inhibited by glutamate receptor antagonists (n=4; 10 µM 6-cyano-7-

nitroquinoxaline-2,3-dione [CNQX] + 50 µM DL-2-amino-5-phosphonovaleric acid [APV]), demonstrating that mEPSCs depict the vesicular release of glutamate. (E). No autoptic synaptic currents were observed after a depolarization step (n=6; -100 mV to +30 mV), indicating that mEPSCs represent interneuronal synaptic transmission. Error bars, SEM.

141

Figure 3.12 - Histogram distribution of mEPSC amplitudes in mouse hippocampal neurons

Amplitude histogram, patch clamp recordings were made from mouse hippocampal neurons before

(A) and after (B) the application of nicotine. Amplitudes of mEPSCs were binned in 10 pA increments, and are represented here as a percentage of the total number of spontaneous synaptic events that occurred during the 10s period analyzed before and after the application of nicotine.

142

3.4.5 MEN1 transcription and menin proteolytic cleavage are disrupted in a mouse model of

Alzheimer’s disease

Tumor suppressor dysfunction has recently been implicated in the synaptic and cognitive deficits

of AD109,110, however a specific link with the cholinergic synaptic loss characteristic of AD

pathophysiology218,219 has not yet been identified. The expression of numerous tumor suppressors,

including MEN1102, are induced by NTF signaling, and reduced levels of NTF expression have

been found in AD228. Considering my observations for the direct actions of menin in nAChR

transcription and functional localization, I next wondered whether menin perturbations might

contribute to cholinergic disruption in AD. To this end, I performed qPCR expression profiling

and WB analysis of menin in hippocampal samples from the 5xFAD transgenic mouse model (n=6

each). The 5xFAD heterozygote (+/-) expresses amyloid precursor protein (APP) and presenilin1

(PS1) mutations associated with familial AD, and recapitulates many features of AD

neuropathology and cognitive deficits beginning at ~ 2 mo229. In the hippocampus of 6 mo

transgenic 5xFAD +/- mice, relative to age-matched littermate controls (5xFAD -/-), I observed a

reduction in the transcript levels of the NTFs BDNF and EGF, and their respective receptors NTRK

and EGFR, MEN1, as well as numerous nAChR subunits, including the MEN1-dependent α5

subunit (Fig. 3.13A; P<0.05-0.001, pair wise fixed reallocation randomization test; Table 3.8).

Expression profiles of the glutamate receptor subunits GluR1 and NR2A and the synaptic marker

Syp, however, did not differ between 5xFAD +/- and -/- mice (P>0.05). At the protein level, I did not observe a significant reduction in the amount of menin in 5xFAD +/- hippocampal samples, despite the transcriptional downregulation of MEN1 (Fig. 3.13B-C; P=0.913, independent t-test;

Table 3.9). However, the relative abundance of the C-menin fragment was significantly reduced

143

(P=0.013), suggesting that mutant APP and PS1 elicit disruptions in the signaling pathways required for calpain-dependent menin proteolytic cleavage. These data suggest that the reduced abundance of the C-menin fragment, which is required for nAChR synaptic consolidation213, may contribute to the dysregulation of cholinergic synapses in AD.

144

Figure 3.13 - MEN1 transcription and menin proteolytic cleavage are disrupted in the

5xFAD mouse model of Alzheimer's disease

145

Figure 3.13: MEN1 transcription and menin proteolytic cleavage are disrupted in the 5xFAD

mouse model of Alzheimer’s disease

(A). Summary data, fold change gene expression in 5xFAD +/- mouse hippocampi, relative to

5xFAD -/- age-matched littermate controls, determined by qPCR (n=6 each, triplicate replicates).

*, statistical significance (pair wise fixed reallocation randomization test), P<0.05-0.001. (B). WB

of 5xFAD -/- and +/- hippocampal protein samples with the C-terminal epitope menin antibody,

depicting full length menin (black arrow) and the C-terminal menin proteolytic fragment (dark grey arrow). Full length menin levels are unchanged, but the C-menin fragment is reduced in

5xFAD +/-. β-actin (βACT) is shown to verify equal loading. (C). Summary data, abundance of the C-menin fragment is reduced in 5xFAD +/-. Asterisk, statistical significance (independent t- test), P<0.05. Error bars, SEM.

146

3.5 Discussion

Whereas the function of menin in cancer biology and transcriptional regulation has been well studied, its role in nervous system function is yet to be fully realized. In the present study, I sought to identify the molecular mechanisms underlying the synaptogenic function of menin in the mammalian CNS. Here, I provide the first evidence that transcriptional regulation of the nAChR

α5 subunit, and the functional clustering of α7 subunit-containing nAChRs involved in presynaptic facilitation, are dependent upon the molecular actions of menin. My data furthermore support an evolutionarily conserved mechanism in which synaptic recruitment of the menin C-terminal proteolytic fragment mediates the functional clustering of nAChRs in neurons, and identify menin perturbations as a potential mechanism contributing to cholinergic synaptic dysfunction in AD.

3.5.1 Menin proteolytic cleavage and differential subcellular distribution

Studies using the menin C-terminal epitope antibody routinely detect a lower band of ~19 kDa in

WBs of protein samples from various mammalian preparations177,230,231, although this has been largely disregarded in the literature as a supposedly non-specific antibody interaction. A recent report from our group, however, characterized the proteolytic cleavage of menin at an evolutionarily conserved calpain consensus site and provided the first indication for a functional significance of the C-terminal menin fragment in mediating the synaptic clustering of neuronal nAChRs213. In the present study, I show that (i) menin in the mouse CNS is similarly cleaved at the evolutionarily conserved calpain site, (ii) menin and the proteolytic fragments exhibit distinct patterns of subcellular localization, and (iii) C-menin co-localizes with α7-nAChRs at

147

glutamatergic presynaptic terminals. These observations suggest that the synaptic clustering of

neuronal nAChRs by menin is a specialized cellular function that has been conserved across evolution. However, calpain expression is ubiquitous and the appearance of this C-menin fragment has been described in multiple tissues, which indicates that the nAChR clustering function of the

C-menin fragment may not be limited to neurons. My observation that the ~19 kDa C-menin immunoreactive fragment is highly prevalent in non-neuronal excitable tissues provide further support for the role of calpain and Ca2+ in the proteolytic cleavage of menin. Menin C-terminal epitope immunoreactivity has been found to localize to the membrane of pancreatic islets232, and

pancreatic β-cells have been reported to express nAChRs233. In line with my present findings, this

localization could be indicative of the C-menin fragment associated with non-neuronal nAChRs,

and suggests that the functional clustering of nAChRs by C-menin may also occur in non-neuronal

tissues.

3.5.2 Menin-dependent transcriptional regulation

Menin is a multifunction transcriptional regulator, acting in complexes with numerous

transcriptional activators, transcriptional repressors, and cell signaling proteins111,178. However,

the impact of menin-dependent transcriptional changes on neuronal and synaptic function is

currently not well understood. We have recently reported that L-menin induces subunit-specific

transcriptional upregulation of nAChRs in Lymnaea neurons213. In the present study, I show that menin knockdown in mouse hippocampal cultures downregulates nAChR α5 transcription, but otherwise induces an apparently nonspecific transcriptional upregulation. The subunit-specific transcriptional regulation of nAChR α5 has two potential explanations: firstly, this may be due to

148

differences in the promoter elements of nAChR genes, with those of α5 binding menin transcriptional activator complexes, whereas other nAChR subunits may have regulatory elements that bind menin transcriptional repressor complexes; secondly, the transcriptional polarity of the nAChR α5 gene is opposite to other nAChR subunits234, and as polarity differentially affects transcriptional activation, this could also be a mechanism for subunit-selective nAChR transcription. Considering the observation reported here that menin expression is restricted to neurons, a previous report that nAChR α5 expression is found in neurons, but not glia235, provides support for the role of menin in subunit-specific transcriptional activation of nAChR α5.

The accumulation of N-menin immunoreactivity alongside the sustained reduction of C-menin upon MEN1 knockdown, as well as the decrease of neuritic C-menin immunoreactivity without a corresponding increase of nuclear N-menin upon calpain inhibition, are intriguing observations that support an autoregulation function for MEN1/menin. Ultimately, these findings may contribute to further insights into the homeostatic function of menin as a tumor suppressor. The accumulation of N-menin over time in culture suggests that the synthesis and stability of menin and the two proteolytic fragments are differentially regulated, likely through an auto-feedback loop involving protein-protein interactions or posttranslational modifications such as activity- or NTF-

dependent phosphorylation204,213. In support of this notion, we have recently reported that the N- terminal L-menin fragment induces transcriptional upregulation of the endogenous L-MEN1 gene, indicating a positive feed-forward system mediated by N-menin213. Conversely, the human MEN1

gene promoter has been found to exhibit decreased activity in response to menin overexpression,

indicating a negative feedback system mediated by menin199. While I show here that the mouse N- menin fragment localizes to the cytoplasm, this does not necessarily exclude the existence of N-

149

menin-dependent transcriptional upregulation of MEN1 in response to reduced intracellular levels of menin, albeit indirectly, as menin translocation to the cytoplasm has been found to shuttle other

transcriptional regulators out of the nucleus180. These findings highlight the complex

transcriptional regulation underlying MEN1 gene induction199, as well as the subsequent gene

targets of menin, whose expression may be regulated by transcription factor complexes that

mediate both transcriptional activation and repression.

3.5.3 nAChRs on presynaptic terminals: subunit composition, function, and targeting

In the CNS, cholinergic synaptic transmission is thought to play a modulatory role in regulating

neuronal network activity236, and the facilitation of neurotransmitter release mediated by nAChR activation has been demonstrated for most classical neurotransmitters in the CNS220. These

observations are consistent with the targeting of nAChRs to presynaptic terminals, and the actions

of nAChRs as ligand-gated, Ca2+-conducting cation channels in increasing the rate of quantal

release through a spatiotemporal calcium-dependent mechanism analogous to paired-pulse facilitation207. The involvement of α-BTX sensitive/α7 subunit-containing nAChRs in nicotine- induced glutamatergic presynaptic facilitation is well documented207,236, however the subunit

composition of native presynaptic nAChRs is currently unknown. The functional and

pharmacological characteristics of presynaptic α7 subunit-containing nAChRs are distinct from those of homomeric α7 channels expressed in heterologous expression systems237, indicating that the native receptors may be non-homomeric α7-containing channels238. Whereas α-BTX blockade

normally eliminates nicotine-induced presynaptic facilitation, knockdown of the α7 subunit results

in the emergence of facilitation that is α-BTX-insensitive, suggesting that either (i) the clustering

150

of non-α7 nAChRs to presynaptic sites occurs as compensation, or (ii) the α7 subunit is one

component of a heteromeric nAChR channel normally involved in presynaptic regulation236. In the

present study, I show that C-menin co-localizes with α7-nAChRs at presynaptic terminals, and that

MEN1 knockdown completely eliminates nicotine-induced presynaptic facilitation. This suggests

that the loss of nAChR-dependent facilitation results from a deficit of nAChR clustering at

presynaptic sites, regardless of nAChR subunit composition. My findings thus raise the intriguing

possibility that the C-terminal menin fragment may be the molecular scaffold required for the

axonal targeting of neuronal nAChRs. The data presented in this study, however, does not allow

me to determine whether this effect is a direct consequence of menin knockdown, or the

consequential knockdown of α5 subunit expression. A previous report that α5-deficient mice

exhibit normal behaviour, brain anatomy, and α-BTX binding patterns239, suggests that presynaptic

nAChR clustering and function occurs normally in the absence of the α5 subunit, and would

therefore preclude a role for α5 in this process. Similarly, α7-deficient mice have also been found

to exhibit normal behaviour and hippocampal-dependent memory240, suggesting that changes to the subunit composition of neuronal nAChR may have relatively little impact on cholinergic synaptic transmission in the CNS. Unfortunately, MEN1-deficient mice exhibit embryonic lethality241, and as such an analysis of the role of menin in brain development and function has not

yet been attempted. Given these current findings, however, this should be pursued with renewed

interest.

At the NMJ, high-density clustering of nAChRs is mediated by the scaffold molecule rapsyn,

which is mobilized by presynaptic agrin signaling through the postsynaptic MuSK receptor

tyrosine kinase76. In contrast to postjunctional muscle membranes, which receive only cholinergic

151

innervation, central neurons receive a variety of synaptic inputs and must meet the additional

challenge of sorting multiple types of neurotransmitter receptors to appropriate synaptic sites. This

feat is usually accomplished by specific molecular interactions and a physical association between

the intracellular domain of a given receptor and a dedicated scaffold or sets of scaffold and adaptor

molecules. Glutamate receptor clustering, for example, is mediated by interactions between

cytoplasmic C-terminal residues either directly with the PDZ domains of PSD-95-type molecular scaffolds (NMDA-type), or indirectly via PDZ domain-containing molecular adaptor proteins

GRIP (glutamate receptor-interacting protein) or PICK1 (protein interacting with C kinase) that bind PSD-95 (AMPA-type)28-30. Similarly, GABA receptor clustering requires the molecular

scaffold gephyrin242, and association of GABA receptors with PDZ scaffolds of the postsynaptic

density occurs by an alternatively spliced isoform of GRIP that binds to gephyrin243. CNS neurons

that are rapsyn deficient, however, exhibit normal synaptic clustering of nAChRs33. PDZ domain-

containing PSD-95 family members have been reported to influence the functional clustering of

various types of neuronal nAChRs, but these molecular scaffolds are also not required244,245. These

observations indicate that distinct, but hitherto undefined molecular scaffolds for neuronal nAChR

clustering are employed in the CNS. My observations are consistent with a role for the C-menin fragment as a scaffold molecule mediating the clustering of α7-nAChRs at presynaptic sites. As pre- and postsynaptic membrane specializations contain multiple elements with PDZ domains, previous reports that PDZ domain-containing scaffold proteins influence the clustering of neuronal nAChRs likely represents a mechanism for the functional localization of nAChRs to appropriate sites adjacent to non-cholinergic presynaptic active zones and postsynaptic densities, given the primarily modulatory role of cholinergic synaptic transmission in the CNS. As menin is not known to contain PDZ domains, and considering its requirement for the functional expression of α7-

152

nAChRs, I suspect that C-menin acts as a dedicated molecular scaffold for particular nAChR

subtypes, and that appropriate synaptic targeting is directed in turn by a yet-to-be determined PDZ

domain-containing molecular adaptor such as PICK1 or GRIP.

3.5.4 The role of menin and nAChRs in synapse formation, plasticity, and maintenance

In the spinal cord dorsal horn, synaptic plasticity is a critical step in the emergence of

hypersensitivity to normally innocuous stimuli following peripheral nerve injury. The

development of neuropathic pain results from a complex suite of functional and anatomical

changes, including enhanced NTF signaling, perturbations to the levels of neurotransmitters and

their receptors, as well as the sprouting of primary afferent fibers and the formation of new

synapses103,104,246. The upregulation of menin105-107 and nAChR α5 subunit8 expression in the

spinal cord dorsal horn have both been reported to mediate the development of neuropathic pain

after peripheral nerve injury. The modulatory α5 subunit is required for the formation of high-

conductance channels that exhibit high ACh sensitivity and calcium permeability6,7, consistent with the promotion of hyper-excitability in neuropathic pain by enhancement of nAChR-facilitated neurotransmitter release. As I found that nAChR α5 subunit gene induction is dependent on MEN1 expression, these observations suggest that the transcriptional actions of menin likely tune nAChR function by regulating subunit composition, and may thus represent a novel transcriptional mechanism underlying the injury-induced development of neuropathic pain or NTF-induced synaptic plasticity.

153

In the hippocampus, activation of nAChRs is known to regulate the expression of NTFs, promote

synapse formation and plasticity, as well as the survival, maturation and synaptic integration of

adult-born neurons247-250. AD pathophysiology is associated with disruption of hippocampal- dependent cognitive processes, a deficit of NTFs, nAChR dysfunction, and the death of cholinergic neurons218,228,251-253. Recently, reduced expression of the BRCA1 and PTEN tumor suppressors

have been implicated in the synaptic dysfunctions underlying AD109,110, contributing to an emerging hypothesis that an interplay between NTF signaling and tumor suppressor dysregulation

underlies the synaptic dysfunction observed in AD. These mechanisms, however, have not yet

been directly linked to the regulation of nAChRs. We have previously shown that mechanisms

underlying L-menin expression, proteolytic cleavage, and synaptic clustering of neuronal nAChRs

are all NTF-dependent102,213. Here, I found reduced levels of NTF/R, MEN1 and numerous nAChR

subunit transcripts, as well as C-menin, but not full menin, in the 5xFAD +/- hippocampus. The reduction of C-menin abundance implicates a shift in the relative distribution of menin towards

the full-length variant, and therefore implicates reduced proteolytic cleavage by calpain. These

findings are intriguing considering that increased intracellular calcium concentrations and hyper- activation of calcium effector proteins, including calpain, are typically associated with AD pathophysiology254,255. How menin escapes proteolytic cleavage in the presence of elevated

calcium and calpain activity remains to be determined, although potentially relevant factors may

include the spatial restriction of calcium transients and calpain activity (e.g. to dendritic arbors or

plaques), the nuclear localization of menin, or posttranslational modifications. Considering that N- menin, but not C-menin immunoreactivity increased over time in response to MEN1 knockdown,

I suspect that the uniform menin protein levels despite transcriptional downregulation of MEN1 in

5xFAD +/- hippocampi reflect compensatory changes underlying menin’s capacity for

154

autoregulation199. Given the cholinergic and synaptic plasticity dysfunctions observed in AD, the

loss of the C-menin fragment, and therefore the functional clustering of α7-nAChR at presynaptic

terminals, may represent a novel mechanism in AD pathophysiology and a new avenue for therapeutic intervention.

155

Table 3.1 - Mouse qPCR gene specific primers (intron spanning)

Target Accession 5’ Sequence 3’ Sequence Efficiency Number (%) β Tubulin NM_023279 AGTCAGCATGAGG TGCAGGTCTGAGT 99.42 GAGATCG CCCCTAC β Actin NM_007393 ACTGTCGAGTCGC GCAGCGATATCGT 101.44 GTCCA CATCCAT MEN1 NM_008583 TCCAGTCCCTCTTC CCAGATGGACATC 96.44 AGCTTC TCTGAGACC nAChR α2 NM_144803 TGAGAAGAATCAA GGAGGTGATGTTG 91.72 ATGATGACCAC CCAAAC nAChR α3 NM_145129 TCAAAGAAGCCAT CACCATGGCAACA 85.10 CCAAAGTG TACTTCC nAChR α4 NM_015730 CGGCCAGTAGCCA AGTCATGCCACTC 90.92 ATATCTC CTGCTTC nAChR α5 NM_176844 TCTGGTTGAAGCA GGATCCACAGAGA 89.37 GGAATGG GTCTGAAGG nAChR α6 NM_021369 GTGGAGAATGTCT CAGCCACAGATTG 103.38 CCGATCC GTCTCC nAChR α7 NM_007390 CCTCTCAGTGGTC AACCATGCACACC 99.83 GTGACAG AATTCAG nAChR β2 NM_009602 ACTCTATGGCGCT GGATCCAAGAGAT 96.71 GCTGTTC GCTCCAC nAChR β3 NM_173212 CCCTGTGTTGAAT GGATTCCATCGTA 91.00 TCCAGTG ATTTTTGG nAChR β4 NM_148944 AGCTCCTCCCAGC AGCCAGATGCTGG 87.76 TCATCTC TGGTC GluR1 NM_008165 ACACAAAGGCCTG ATGGCTTGGAGAA 98.13 GAATGG GTCGATG NR2A NM_008170 TCATGATCCAGGA ATCGGAAAGGCGG 95.35 GGAGTTTG AGAATAG SYP NM_009305 GAGGGACCCTGTG AGCCTGTCTCCTT 92.11 ACTTCAG GAACACG BDNF NM_007540 GGGTTAACTTTGG TGGTCATCACTCTT 93.54 GAAATGC CTCACCTG NTRK NM_001025074 CATGGTCTTTGAG AGCTCTGTGGGCG 92.06 TACATGAAGC GGTTAC EGF NM_010113 GGATGGTACGAAT TTCCATCTATGTG 90.58 GGTGCAG GGGCTTC EGFR NM_207655 ACACTGCTGGTGT TCCTCTGCAGGCT 91.42 TGCTGAC CAGAAAG

156

Table 3.2 - Immunocytochemistry fluorescence of menin in hippocampal neuron cultures

Treatment Nuclear N-menin Mean Fluorescence P 1 n 2 ± SEM (AU) 0.1% DMSO 86.66 ± 3.83 56 0.238 20 μM PD150606 92.52 ± 2.98 50 Treatment Cytoplasmic C-menin Mean Fluorescence P n 3 ± SEM (AU) 0.1% DMSO 34.74 ± 1.52 33 <0.001 20 μM PD150606 24.33 ± 1.16 30

1. Independent samples t-test (2-sided)

2. n refers to the analysis of N-menin epitope fluorescence intensity within neuronal nuclei

(delineated by DAPI), 10-11 separate images from 3 independent experiments

3. n refers to the analysis of C-menin epitope fluorescence intensity within a 20 μM length

of neurite, 3 separate neurites per image were analyzed, 10-11 separate images from 3

independent experiments

Table 3.3 - Incidence of C-menin co-localization in super-resolved immunocytochemistry images

ICC Label Mean Number of % Co-localization P 1 n 2 Puncta ± SEM with C-menin ± SEM Synaptotagmin 43.89 ± 5.72 61.02 ± 7.53 0.006 9 PSD-95 46.00 ± 3.55 29.11 ± 2.87 <0.001 10 α-BTX 45.79 ± 4.19 93.40 ± 2.01 (F=78.187; P<0.001) 19

1. One-way ANOVA with Games-Howell post hoc test, significance (P) relative to C-menin

and α-BTX co-localization

2. n refers to ≥9 separate images from ≥2 independent experiments

157

Table 3.4 - Relative gene expression in hippocampal neuron cultures

Treatment Fold-Change qPCR Target Standard Error P 2 Expression 1 Untreated MEN1 1 0.847 - 1.181 - Control nAChR α 2 1 0.803 - 1.245 - nAChR α 3 1 0.610 - 1.641 - n=6 nAChR α 4 1 0.863 - 1.159 - nAChR α 5 1 0.722 - 1.386 - nAChR α 6 ND 3 - - nAChR α 7 1 0.781 - 1.280 - nAChR β 2 1 0.851 - 1.175 - nAChR β 3 ND - - nAChR β 4 1 0.377 - 2.650 - GluR1 1 0.908 - 1.101 - NR2A 1 0.847 - 1.181 - SYP 1 0.863 - 1.158 - NTC shRNA MEN1 0.812 0.649 - 1.179 0.195 nAChR α 2 1.212 0.925 - 1.609 0.148 n=6 nAChR α 3 1.181 0.683 - 1.907 0.394 nAChR α 4 1.213 0.955 - 1.463 0.059 nAChR α 5 1.315 0.935 - 1.831 0.098 nAChR α 6 ND - - nAChR α 7 1.002 0.713 - 1.365 0.982 nAChR β 2 1.564 1.117 - 2.340 0.014 nAChR β 3 ND - - nAChR β 4 0.795 0.324 - 1.783 0.522 GluR1 1.203 1.010 - 1.496 0.061 NR2A 0.651 0.449 - 0.971 0.019 SYP 1.174 0.929 - 1.551 0.165 MEN1 shRNA MEN1 0.593 0.430 - 0.933 0.009 nAChR α 2 2.964 2.200 - 3.910 0.001 n=6 nAChR α 3 2.983 1.661 - 5.317 0.001 nAChR α 4 1.569 1.059 - 2.198 0.019 nAChR α 5 0.343 0.256 - 0.446 0.002 nAChR α 6 ND - - nAChR α 7 2.196 1.553 - 3.183 0.001 nAChR β 2 3.728 2.762 - 5.242 0.001 nAChR β 3 ND - - nAChR β 4 1.392 0.515 - 3.080 0.383 GluR1 2.658 2.184 - 3.274 0.001 NR2A 3.729 2.476 - 6.935 <0.001 SYP 2.727 2.220 - 3.325 0.001

158

1. Relative expression, normalized to β-Actin and β-Tubulin reference genes

2. Pair wise fixed reallocation randomization test (REST-2009), significance (P) relative to

untreated control

3. ND indicates the transcript was not detected in the qPCR reaction

Table 3.5 - Immunocytochemistry fluorescence of menin and nAChRs in NTC shRNA- transduced hippocampal neuron cultures

Treatment DIV Menin (C) P 1 n 2 Menin (N) P n NTC shRNA Fluorescence Fluorescence ± SEM (AU) ± SEM (AU)

GFP+ 3 59.64 ± 5.83 0.536 28 71.66 ± 4.42 0.976 36 GFP- 55.31 ± 4.15 42 71.85 ± 4.63 27

GFP+ 7 76.03 ± 4.28 0.515 33 71.76 ± 3.30 0.970 49 GFP- 72.28 ± 3.85 38 71.95 ± 3.91 33

GFP+ 10 56.91 ± 3.24 0.745 29 96.67 ± 3.50 0.554 32 GFP- 58.20 ± 2.39 34 100.68 ± 5.74 30

GFP+ 14 52.93 ± 3.49 0.911 25 98.58 ± 4.54 0.877 43 GFP- 52.48 ± 2.26 30 97.59 ± 4.48 46 Treatment DIV nAChR α5 P n α-BTX P n NTC shRNA Fluorescence Fluorescence ± ± SEM (AU) SEM (AU)

GFP+ 3 49.47 ± 3.77 0.859 27 72.67 ± 4.54 0.588 45 GFP- 50.75 ± 6.70 18 69.13 ± 4.61 52

GFP+ 7 46.51 ± 2.31 0.634 24 103.29 ± 5.72 0.586 53 GFP- 48.14 ± 2.41 30 99.22 ± 4.81 54

GFP+ 10 50.27 ± 2.07 0.682 23 117.11 ± 5.00 0.768 48 GFP- 49.05 ± 2.11 24 114.93 ± 5.35 51

GFP+ 14 46.84 ± 2.76 0.968 30 50.27 ± 2.07 0.682 23 GFP- 46.70 ± 2.27 24 49.05 ± 2.11 24

1. Independent samples t-test (2-sided)

2. n refers to the analysis of fluorescence intensity within neuronal soma, ≥4 separate

images from ≥2 independent experiments

159

Table 3.6 - Immunocytochemistry fluorescence of menin and nAChRs in MEN1 shRNA- transduced hippocampal neuron cultures

Treatment DIV Menin (C) P 1 n 2 Menin (N) P n MEN1 Fluorescence Fluorescence shRNA ± SEM (AU) ± SEM (AU)

GFP+ 3 40.12 ± 2.42 <0.001 28 60.07 ± 2.88 0.005 54 GFP- 64.16 ± 3.34 37 74.58 ± 4.07 45

GFP+ 7 43.85 ± 1.69 <0.001 50 65.38 ± 2.72 0.001 58 GFP- 60.56 ± 3.44 26 79.71 ± 3.40 44

GFP+ 10 53.02 ± 3.32 0.003 24 83.77 ± 4.77 0.569 47 GFP- 70.88 ± 4.89 17 79.41 ± 5.65 27

GFP+ 14 43.52 ± 3.23 0.005 14 56.86 ± 3.96 0.815 35 GFP- 62.60 ± 5.42 13 55.31 ± 5.53 27 Treatment DIV nAChR α5 P n α-BTX P n Fluorescence ± Fluorescence ± SEM (AU) SEM (AU)

GFP+ 3 40.21 ± 3.76 0.731 24 75.67 ± 4.26 0.024 54 GFP- 42.20 ± 4.32 25 90.06 ± 4.58 61

GFP+ 7 41.10 ± 2.19 <0.001 36 111.46 ± 3.17 0.456 86 GFP- 57.63 ± 2.63 29 115.31 ± 4.02 51

GFP+ 10 42.35 ± 3.13 0.001 18 96.56 ± 4.14 0.682 58 GFP- 59.91 ± 3.32 22 99.23 ± 4.85 36

GFP+ 14 41.37 ± 2.72 <0.001 12 97.30 ± 10.14 0.186 14 GFP- 58.96 ± 2.90 23 115.60 ± 8.71 13

1. Independent samples t-test (2-sided)

2. n refers to the analysis of fluorescence intensity within neuronal soma, ≥4 separate

images from ≥2 independent experiments

160

Table 3.7 - mEPSC amplitude and frequency values in hippocampal neuron cultures

Treatment Mean Standard P 1 P 2 n Amplitude Error (pA) (pA) Untreated Control 33.09 2.88 (F=0.405; P=0.844) Pre-Nicotine 0.800 19 Untreated Control 32.55 2.48 Post-Nicotine 1.000 NTC shRNA 37.69 2.86 Pre-Nicotine 0.912 0.259 15 NTC shRNA 35.95 2.60 Post-Nicotine 0.988 MEN1 shRNA 36.33 3.99 Pre-Nicotine 0.976 0.717 17 MEN1 shRNA 35.92 3.89 Post-Nicotine 0.987 Treatment Mean Standard P P n Frequency Error (Hz) (Hz) Untreated Control 3.60 0.58 (F=1.373; P=0.241) Pre-Nicotine <0.001 19 Untreated Control 4.87 0.67 Post-Nicotine 0.494 NTC shRNA 3.66 0.50 Pre-Nicotine 1.000 <0.001 15 NTC shRNA 4.86 0.55 Post-Nicotine 0.571 MEN1 shRNA 3.74 0.37 Pre-Nicotine 1.000 0.628 17 MEN1 shRNA 3.61 0.43 Post-Nicotine 1.000

1. One-way ANOVA with Tukey’s HSD post hoc test, significance (P) relative to untreated

control pre-nicotine

2. Paired samples t-test (2-sided)

161

Table 3.8 - Gene expression in 5xFAD +/- relative to 5xFAD -/- hippocampus

Genotype Fold-Change qPCR Target Standard Error P 2 Expression 1 BDNF 0.352 0.249 - 0.551 0.006 5xFAD +/- NTRK 0.302 0.233 - 0.425 0.003 (n=6) EGF 0.303 0.190 - 0.497 0.004 EGFR 0.484 0.310 - 0.728 0.006 Relative to MEN1 0.516 0.398 - 0.709 0.004 nAChR α 2 0.405 0.190 - 0.799 0.021 5xFAD -/- nAChR α 3 0.458 0.207 - 1.152 0.072 (n=6) nAChR α 4 0.335 0.238 - 0.457 0.001 nAChR α 5 0.309 0.215 - 0.451 0.002 nAChR α 6 0.492 0.243 - 1.264 0.070 nAChR α 7 0.529 0.376 - 0.764 0.002 nAChR β 2 0.468 0.371 - 0.591 <0.001 nAChR β 3 0.422 0.224 - 0.812 0.010 nAChR β 4 0.567 0.325 - 1.233 0.076 GluR1 0.994 0.528 - 1.544 0.986 NR2A 0.926 0.550 - 1.372 0.721 Synaptophysin 1.009 0.806 - 1.298 0.934

1. Relative expression, normalized to β-Actin and β-Tubulin reference genes

2. Pair wise fixed reallocation randomization test (REST-2009)

Table 3.9 - Western blot analysis of menin in 5xFAD +/- and 5xFAD -/- hippocampus

Genotype Menin (C) C-Menin (C) Relative n P 2 Fluorescence Fluorescence Fluorescence ± SEM (AU) 1 ± SEM (AU) 1 C-Menin / Menin

5xFAD -/- 1.40 ± 0.05 1.25 ± 0.11 0.94 ± 0.03 6 0.013 5xFAD +/- 1.42 ± 0.02 1.15 ± 0.03 0.79 ± 0.04 6

1. Normalized to the fluorescence value of the loading control β-Actin

2. Relative fluorescence, independent samples t-test (2-sided)

162

Chapter Four: A novel mechanism of synaptic specificity: synaptophysin inhibits

presynaptic secretory machinery to regulate the release of peptide neurotransmitters

Sections of this chapter are being prepared for publication in the following manuscript:

Getz A, Janes TA, Visser F, Zaidi W, Deshwar A, Kawasoe J & Syed NI (2016). A novel mechanism of synaptic specificity: synaptophysin inhibits presynaptic secretory machinery to regulate the release of peptide neurotransmitters.

4.1 Abstract

A presynaptic neuron can concurrently and/or differentially release a multitude of

neurotransmitters from its individual presynaptic terminals to selectively modulate the activity of

distinct postsynaptic targets within a neuronal network. Neuropeptides are important signals for

synaptic plasticity that can be selectively employed, however, the mechanisms that regulate their

release remain undefined. Here, I demonstrate that a co-transmitting Lymnaea neuron forms either

a purely cholinergic synapse or a mixed cholinergic-peptidergic synapse with two distinct

postsynaptic targets. My results identify a novel role for synaptophysin as an effector of synapse-

specific neuropeptide release, via the inhibitory regulation of large dense-core vesicle secretory

machinery. This mechanism of presynaptic transmitter specificity is dynamically regulated by both

extrinsic neurotrophic factors and target cell-specific retrograde arachidonic acid signaling.

Moreover, it is established in the presence of competing molecular signals from distinct postsynaptic targets both in vitro and in vivo. These findings define a novel molecular mechanism for synaptic specificity and plasticity through which the co-transmitter characteristics of individual synapses can be tuned to meet the spatiotemporal functional requirements of neuronal networks.

163

4.2 Introduction

Most, if not all neurons use more than one type of neurotransmitter to mediate synaptic

transmission3,256. The use of multiple transmitters by a neuron enhances its capacity to encode

information, and the selective use of co-transmitter substances24,138 underlies the functional

flexibility of synaptic networks. Many neurons form hundreds of presynaptic terminals, innervate

a wide variety of postsynaptic targets, and modulate their output (i.e. release of neurotransmitters)

in response to intracellular and extracellular signals11. Whereas the initial theory of synaptic

transmission suggested that neurons, as a single metabolic unit, would store and release the same

transmitters at all of their synaptic terminals (Dale’s Principle257), the functional specificity of individual presynaptic terminals is becoming increasingly accepted as a universal phenomenon.

There is accumulating evidence that the molecular composition, co-transmitter localization and release characteristics of a neuron’s individual presynaptic terminals are specified according to the identity of the postsynaptic target cell126,128-134128,135, and the possibility that this might be modified

by synaptic plasticity events is also emerging127,134. However, the molecular mechanisms underlying this form of target cell-dependent presynaptic specificity and plasticity remain largely unidentified. Consequently, exactly how a neuron discretely tunes the composition and function

of its presynaptic terminals, necessary for the specificity of synaptic networks and to meet an

animal’s demands for behavioural flexibility, remains a fundamental unanswered question.

Neuropeptides are selectively employed co-transmitter substances that regulate numerous aspects

of synaptic plasticity, physiology and behaviour258. However, the mechanisms controlling the

release of peptide neurotransmitters from presynaptic terminals have remained largely undefined

164

because of the technical challenges associated with accessing and manipulating peptidergic

synapses, or defining the precise nature of synaptic transmission and molecular signaling

interactions between pre- and postsynaptic neurons. Several lines of evidence indicate that

invertebrate co-transmitting neurons vary the use of classical and peptide neurotransmitters at

synapses with different postsynaptic targets, and this specificity cannot be explained by

postsynaptic variations in receptor expression137-139,259. Furthermore, although neuropeptides are

most commonly viewed as slow-acting neuromodulators released from active zone-adjacent sites

in response to bursting activity260, they have also been shown to cluster directly at presynaptic active zones and serve as fast-acting mediators of synaptic transmission in response to a single action potential138,154. These observations suggest that the peptidergic characteristics of

presynaptic terminals can be highly variable, although the molecular mechanisms that generate

such differences are yet to be determined.

To address these question, I employed identified neurons from the central nervous system (CNS)

of the invertebrate mollusk Lymnaea stagnalis, and studied the neurotransmitter release

characteristics of an identified co-transmitting neuron at both in vitro recapitulated and in vivo

synapses with two distinct postsynaptic targets. I found that synaptophysin (Syp), a synaptic

vesicle protein (SVP) of ambiguous function, defined the neuropeptide release competency of a

synapse. Synaptophysin expression was regulated by both extrinsic neurotrophic factors (NTFs) and trans-synaptic retrograde signaling induced by the target cell-specific generation of arachidonic acid metabolites (AA). My results (i) demonstrate a novel role for Syp in the inhibitory regulation of peptide neurotransmitter release, (ii) delineate a target cell-specific retrograde signaling mechanism underlying the functional differentiation of individual presynaptic terminals,

165

and (iii) define a new form of NTF-dependent synaptic plasticity through which the co-transmitter characteristics of individual presynaptic terminals can be regulated.

166

4.3 Methods

4.3.1 Animals and neuronal cell culture

Lymnaea stagnalis were raised under standard conditions in freshwater aquaria at room

temperature (~22°C). Animals (6-8 weeks old) were anesthetized with a 10% Listerine solution

(10m) and sacrificed by dissection of the CNS. Neurons were isolated from trypsinized CNS by

suction applied through a glass pipette, and cultured on poly-L-lysine coated glass culture dishes, as previously described in detail151. Neurons were maintained overnight (15-24h) in defined media

(DM; L-15 [Life technologies]) or CM (CNS-incubated DM, generates NTF-rich media).

4.3.2 Molecular biology

Synaptotagmin I C2B-α (Syt-α), C2B-β (Syt-β), and Syp were cloned from Lymnaea CNS cDNA,

and eGFP from the pWPI lentiviral construct (Addgene), as described elsewhere102. Primers are

shown in Table 4.1. Synthetic mRNA was made with the mMESSAGE mMACHINE T7 Ultra

transcription kit (Ambion). mRNA was microinjected into VD4 neurons with a sterile low

resistance glass electrode (10 pulses, 250 ms, 10 PSI), and molecular-grade water was

microinjected as a vehicle control. Quantitative (q)PCR expression profiling of cultured neurons,

including negative controls and validations, was performed as previously described in detail102.

Expression profiling of in situ neurons was performed with the following modifications: identified neurons were isolated from the CNS as above, then deposited into PCR reaction buffer and immediately frozen on dry ice; cells were lysed at 75°C for 10m, and DNA was digested with

167

DNase (Qiagen). Gene-specific primers and efficiency values are shown in Tables 4.2-4.3.

-ΔC Expression (2 T) was normalized to 18s ribosomal (r)RNA.

4.3.3 Immunocytochemistry, immunohistochemistry, in situ hybridization and imaging

Immunocytochemistry: Neurons were fixed for 30m with 1% paraformaldehyde (PFA) and 0.2%

picric acid in 1xPBS, then permeabilized for 1h with incubation media (IM) containing 0.5%

Triton and 10% goat serum in 1xPBS. The primary antibody against FMRFamide neuropeptides

183 (FMRFa), which detects the Arg-Phe-NH2 moiety , was used at 1:500 in IM for 1h. Secondary

antibodies (Alexa Fluor 488 or 546 conjugated [Invitrogen]) were used at 1:100 in IM for 1h.

Three 15m 1xPBS washes were performed between each incubation, and all incubations were

performed at room temperature. Cells were mounted with MOWIOL containing DAPI.

Immunohistochemistry & in situ hybridization: CNS were dissected and fixed in 4% PFA for 18-

24h, incubated in 30% sucrose for 24h, then frozen in OCT embedding medium (Tissue-Tek) at -

80°C. 8 μm cryostat sections were used for immunohistochemistry (IHC), as above for

immunocytochemistry (ICC), or in situ hybridization (ISH), as previously described261.

Digoxygenin-labeled sense and antisense probes were synthesized using the DIG-RNA labeling kit and T7 or T3 RNA polymerase (Roche) from linearized pBluescript plasmids (Addgene) containing the target sequences. Probes are shown in Table 4.4.

Samples were imaged with an LSM 510 Meta microscope using a Plan-Apochromat 20x/0.75na objective (Zeiss), or an A1R MP microscope using a CFI Plan Fluor 20x/0.75 MI objective

(Nikon). Laser excitation wavelengths were 402, 488 and 561, in series, and emission wavelengths

168

were collected through a 560-615 nm bandpass filter (Zeiss) or 450/50, 525/50 and 595/50 filter

cubes (Nikon). For quantitative analysis, imaging parameters were kept the same amongst relevant

samples. Images were acquired with LSM 510 software (Zeiss) or NIS Elements v4.13.00 (Nikon),

and intensity analysis performed with ImageJ (NIH).

4.3.4 Electrophysiology

Intracellular current clamp recordings and Lucifer yellow dye injections were performed as

previously described for in vitro101 and in situ262 preparations. To determine acetylcholine (ACh)

release characteristics, the mean amplitude of 5 consecutive postsynaptic potentials (PSPs), at a

holding potential of -100 mV, in visceral F group (VF) or right pedal dorsal 1 (RPeD1) neurons

was measured in response to single action potentials elicited in visceral dorsal 4 (VD4). To

determine FMRFa release characteristics, the of excitation of VF at holding potentials of -60 mV

(in vitro) or -70 mV (in situ), or inhibition of RPeD1 at -50 mV (in vitro) or -60 mV (in situ), was

measured in response to a ~ 4s burst elicited in VD4. Peptidergic transmission was isolated

pharmacologically by an ACh receptor (AChR) antagonist cocktail: 5 μM methyllcaconitine

(MLA); 10 μM tubocurarine (TC); 20 μM tetraethylammonium (TEA); incubation time 5-10m157.

4.3.5 Chemicals

FMRFa (Phe-Met-Arg-Phe-NH2), ACh, MLA, TC, TEA and pertussis toxin (PTX) were

purchased from Sigma-Aldrich. NG-Nitro-L-arginine (L-NNA), gallein and AA were purchased from Tocris. 5,8,11,14-Eicosatetraynoic acid (ETYA) was purchased from Santa Cruz Biotech.

0.1% DMSO and Tocrisolve 100 (Tocris) vehicle controls were performed for all experiments.

169

4.3.6 Experimental design and statistical analysis

All reported results are derived from ≥2 independent experiments, to ensure reliability and replicability. Sample sizes were limited by uncontrollable factors inherent to the cell culturing techniques used. Data analysis was performed blinded by acquisition file number. Statistical analyses were performed using SPSS Statistics v22. Peptidergic synaptic frequency was assessed with Pearson’s Chi-squared test (2-sided). ICC fluorescence intensity and synaptic physiology were assessed with Student’s independent samples t-test (2-sided) or univariate ANOVA with either Tukey’s HSD (equal variance; Levene statistic P>0.05) or Games-Howell (unequal variance; Levene statistic P<0.05) post hoc tests, as appropriate.

170

4.4 Results

4.4.1 Postsynaptic target cell specificity and extrinsic neurotrophic factors regulate peptidergic

synaptic transmission

I first sought to develop a synaptic assay with which I could detect the differential release of

classical and peptide transmitters from a single presynaptic neuron. The Lymnaea VD4 (Visceral

dorsal 4) interneuron expresses the classical small molecule transmitter ACh alongside heptapeptides derived from FMRFamide gene transcripts (G/SDPFLRF-NH2; discussed in text as

FMRFa neuropeptides for simplicity), and forms monosynaptic connections with many identified

neurons in the CNS162,187,263. In different cell populations FMRFa neuropeptides exert opposite

membrane actions (excitatory vs. inhibitory)9. These cell-specific actions have been attributed to

coupling of the FMRFa G-protein-coupled receptor (GPCR) to distinct G-protein complexes264.

As GPCRs can activate a range of metabolic signaling pathways, I co-cultured VD4 with

postsynaptic targets that exhibit opposite responses to FMRFa, reasoning that differences in trans- synaptic cell-cell signaling might underlie the emergence of presynaptic transmitter specificity

(Fig. 4.1H; n=3 each). To this end, VD4 was co-cultured with either the postsynaptic neuron VF

(Visceral F group) or RPeD1 (Right pedal dorsal 1). I made dual intracellular recordings from soma-soma paired neurons, first under control conditions, and then in the presence of an AChR antagonist cocktail (5 μM MLA, 10 μM TC, 20 μM TEA157) to isolate the peptidergic component

of synaptic transmission (Fig. 4.1-4.2). When cultured in CNS-conditioned media (CM; NTF- rich), which resembles the in vivo environment, VD4-VF formed a mixed cholinergic-peptidergic synapse, where all VD4 neurons released both ACh and FMRFa (Fig. 4.1A; n=8/8 peptidergic;

171

Table 4.5). VD4-RPeD1 formed a purely cholinergic synapse, where all VD4 neurons released

ACh, but few released FMRFa (Fig. 4.1B; n=3/10 peptidergic; P=0.012, Chi-squared test).

As NTF signaling governs many aspects of synapse formation, maturation, and plasticity at the presynaptic terminal71, I next cultured VD4-VF and VD4-RPeD1 neurons in the absence of NTF

(DM; defined media, contains no NTF) to determine whether synapse-specific peptidergic transmission might be regulated by extracellular factors. Surprisingly, I observed a complete reversal of peptidergic phenotype – VD4-VF formed a purely cholinergic synapse, where few VD4 neurons released FMRFa (Fig. 4.1C; n=2/10 peptidergic), whereas VD4-RPeD1 formed a mixed cholinergic-peptidergic synapse, where most VD4 neurons released both ACh and FMRFa (Fig.

4.1D; n=9/12 peptidergic; P=0.010, Chi-squared test). These data suggest that NTF signaling and target cell identity collectively influence presynaptic transmitter specificity.

172

Figure 4.1 – Presynaptic transmitter specificity is regulated by neurotrophic factors and

postsynaptic target identity

173

Figure 4.1: Presynaptic transmitter specificity is regulated by neurotrophic factors and

postsynaptic target identity

(A-D). Synaptic assay for the differential release of neuropeptides and the characterization of

synaptic transmission by intracellular current clamp recordings. A mixed cholinergic-peptidergic synapse forms between VD4-VF in CM (A), and VD4-RPeD1 in DM (D). A purely cholinergic synapse forms between VD4-RPeD1 in CM (B), and VD4-VF in DM (C), demonstrating the modulation of peptide transmitter specificity by NTF and postsynaptic identity. N≥8. Inserts show

ACh-mediated PSPs (left; -100 mV holding potential), inhibited by AChR antagonists (right; 5

μM MLA, 10 μM TC, 20 μM TEA). (E-F). ICC labeling of FMRFa neuropeptides in VD4-VF (E) and VD4-RPeD1 (F) soma-soma pairs cultured in CM. N≥7. Scale bars, 20 μm. (G). Summary data, incidence of synapses exhibiting peptidergic transmission. *, P<0.05; ** P<0.01 (Chi- squared test). (H). The excitatory peptidergic postsynaptic response of VF is unaffected by 100

μM amiloride (FMRFa-gated sodium channel inhibitor; left), but is inhibited by 100 ng/mL PTX

(Gαi/o inhibitor; middle). The inhibitory peptidergic postsynaptic response of RPeD1 is unaffected

by PTX (right), indicating that the FMRFa receptor is differentially coupled to G proteins in these

two postsynaptic targets. N=3 each.

174

Figure 4.2 - Postsynaptic receptor expression profiles

175

Figure 4.2: Postsynaptic receptor expression profiles

(A). VD4-VF were cultured in a soma-soma configuration in CM. Pressure application (arrows;

250 ms pulse, 10 PSI, tip opening ~ 1-5 μm) of 1 μM ACh elicits an inhibitory response in VF (i),

which is eliminated in the presence of the AChR antagonist cocktail (ii; 5 μM MLA, 10 μM TC;

20 μM TEA). (B). FMRFamide heptapeptides and tetrapeptides elicit opposite responses in the

same cell, and FMRFamide heptapeptides or tetrapeptides can also elicit opposite responses in

different cells9, through cell-specific coupling of the receptors to different G protein complexes264.

As FMRFamide heptapeptides are not commercially available, I used the FMRFamide tetrapeptide

(Phe-Met-Arg-Phe-NH2) to evaluate the postsynaptic responses of VF and RPeD1 to FMRFa

neuropeptides. VF and RPeD1 were paired in a soma-soma configuration with VD4 and cultured

in CM. Pressure application of 1 mM FMRFa tetrapeptide (arrows; 250 ms pulse, 10 PSI, tip

opening ~ 1-5 μm) elicits inhibition in VF neurons (i) and excitation in RPeD1 neurons (ii). Note

that these responses to the FMRFamide tetrapeptide are opposite of the postsynaptic responses to

the FMRFa heptapeptides released by VD4162 (G/SDPFLRFamide; VF, excitatory; RPeD1,

inhibitory; see Fig 4.1). (C). NTFs do not change the profiles of postsynaptic responses to FMRFa

neuropeptides. Pressure application of 1 mM FMRFamide tetrapeptide (arrows; 250 ms pulse, 10

PSI, tip opening ~ 1-5 μm) also elicits inhibition in VF cultured with VD4 in DM (i). The response

to FMRFa is unaffected by the AChR antagonist cocktail (ii). N≥3 each.

176

4.4.2 Presynaptic inhibition of neuropeptide secretory machinery defines NTF- and target cell-

dependent presynaptic transmitter specificity

Considering that postsynaptic receptor expression profiles do not account for target cell-specific absences of peptidergic transmission138,139, there exist three alternative scenarios to explain the

presynaptic physiology: (i) the expression of neuropeptides changes; (ii) neuropeptides are not

targeted to presynaptic sites; or (iii) neuropeptides are not released from presynaptic sites. To

explore these possibilities, I used α-FMRFa ICC, and neurons were cultured in an axon-axon

configuration to facilitate visualization of synaptic sites. Intensity of the somatic FMRFa

immunolabel was equal amongst unpaired VD4, VF-paired VD4, and RPeD1-paired VD4 cultured

in CM or DM, indicating that the expression of FMRFa neuropeptides does not change in response

to target cell contact or NTF (Fig. 4.3A-G; n≥7; P≥0.967, one-way ANOVA; Table 4.6). Neuron

pairs that form purely cholinergic synapses were characterized by a ‘hyper-innervation’

phenotype, in which intensity of the FMRFa immunolabel at presumptive synaptic sites was

significantly increased relative to the pairs that form mixed cholinergic-peptidergic synapses (Fig.

4.3H; P≤0.037, one-way ANOVA; Table 4.7). These results indicate that neuropeptides are being

targeted to, but not released from the presynaptic sites that exhibit purely cholinergic transmission.

177

Figure 4.3 - Inhibition of neuropeptide release defines presynaptic transmitter specificity

178

Figure 4.3: Inhibition of neuropeptide release defines presynaptic transmitter specificity

(A-F). ICC labeling of FMRFa neuropeptides in unpaired VD4 (A,B) VF-paired VD4 (C,D) and

RPeD1-paired VD4 (E,F), cultured in CM (A,C,E) or DM (B,D,F). Arrowheads, putative peptide releasing synaptic sites; arrows, putative peptide non-releasing synaptic sites (according to the context-dependent peptidergic synaptic phenotype observed in Fig. 4.1). N≥7. Scale bars, 50 μm.

(G). Summary data, the somatic fluorescence intensity is unaffected by NTF and target cell contact

(P≥0.967, one-way ANOVA). (H). Summary data, fluorescence intensity is increased at synapses

that do not exhibit peptidergic transmission. *, P<0.05-0.005 (one-way ANOVA). Error bars,

SEM.

179

4.4.3 Presynaptic expression of synaptophysin is regulated by neurotrophic factors and

modulated by postsynaptic identity

The context-dependent accumulation of FMRFa neuropeptides indicates that large dense-core

vesicle (LDCV) release varies according to both the identity of the postsynaptic neuron and NTF

signaling. Since LDCVs do not undergo recycling, changes in gene expression should be key to

influencing the SVP composition, and therefore the release competency of LDCVs. The mammalian synaptotagmin (Syt) IV isoform is known to negatively regulate LDCV fusion173 and

is an immediate early gene induced by neuronal activity175, making variations in synaptotagmin

isoform expression a candidate mechanism for presynaptic inhibition. Another potential negative

regulator of release is the synaptobrevin-synaptophysin (Syb; Syp) complex. The v-SNARE Syb, when bound to the functionally undefined integral membrane protein Syp, is unable to form the fusogenic SNARE complex18,265,266. I sought to use a candidate molecule single-cell qPCR approach to identify SVP gene expression differences induced by NTF and target cell contact. The sequences for several Lymnaea SVPs, including Syt I, Syb and syntaxin (Syx) have been previously identified. I searched the Lymnaea expressed sequence tag (EST) library150 and found

a single Syp-like sequence that is homologous to mammalian Syp and exhibits the characteristic 4 transmembrane domains of Syp family proteins (Fig. 4.4; Accession no. ES572211). I cloned

Lymnaea Syp and identified 2 splice variants, resulting in a truncation of 8 C-terminal residues, which I designate Syp-L (long) and Syp-S (short) (Fig. 4.5). In the marine mollusc Aplysia, Syt I

splice variants, C2B-α and C2B-β, have been found to have different affinities for v-SNARE

binding267, and the C2B-α isoform has been reported to act as an inhibitory clamp on synaptic

vesicle (SV) release268, suggesting that invertebrate mollusks have a system for facilitative and

180

inhibitory regulation of SVs that might be similar to the one present in mammals174. The previously

identified Lymnaea Syt I sequence (Accession no. AF484090) is highly homologous to the C2B-

β Aplysia isoform, so I cloned Syt I from the Lymnaea CNS and screened clones for the C2B-α

and C2B-β splice variants. I identified a novel Lymnaea Syt I C2B-α-like isoform with ~50%

prevalence, suggesting that it is highly expressed in the CNS (Figs. 4.6-4.7). The Lymnaea Syt I

C2B-α and C2B-β isoforms have 9 amino acid substitutions resulting from alternative exon usage,

which is an identical scenario to the two Aplysia isoforms267 (Fig. 4.6). Therefore, I propose that

the Lymnaea Syt I isoforms be similarly designated Syt I C2B-α and C2B-β (Syt-α; Syt-β).

I then used this SVP candidate molecule approach and single cell qPCR profiling to screen for differential gene expression in unpaired, VF-paired, and RPeD1-paired VD4 neurons cultured in

CM or DM (Fig. 4.8A-C; n=7-10 neurons, 3 qPCR triplicate replicates; Tables 4.8-4.9). I also characterized the expression of FMRFamide heptapeptide and tetrapeptide transcripts as an additional validation for FMRFa expression, and screened G protein expression profiles of VF and

RPeD1 to determine whether changes in metabolic signaling amongst the two cell types or in different culture conditions might underlie the emergence of peptide transmitter specificity (Fig.

4.8D-E; Tables 4.10-4.11). While a number of context-dependent variations were observed, only the expression profiles of Syp exhibited a prominent reversal pattern, which indicates that low expression levels in VD4 may facilitate LDCV release, whereas high expression levels might inhibit LDCV release. Unpaired VD4 neurons exhibited similar Syp profiles as RPeD1-paired

VD4 in both CM and DM, suggesting that (i) NTFs influence Syp expression and (ii) molecular signals from VF counteract this, resulting in NTF- and target cell-dependent regulation of Syp

181

expression. These observations identify a putative role for Syp in negatively regulating the release of LDCVs.

182

Figure 4.4 - Prediction of Lymnaea synaptophysin transmembrane domains

183

Figure 4.4: Prediction of Lymnaea synaptophysin transmembrane domains

(A-B). Amino acid residue hydrophobicity plots predict the membrane topology and location of transmembrane helices in a Lymnaea Syp-like protein (A; accession no. ES572211) and mouse

Syp (B; accession no. NM009305). Blue indicates predicted cytoplasmic residues, red indicates predicted transmembrane residues, pink indicates predicted intravesicular residues. (TMHMM 2.0;

CBS). (C). Protein sequence alignment of Lymnaea (top) and mouse (bottom) Syp (Clustal Omega;

EMBL-EBI). Grey bars indicate predicted Lymnaea transmembrane domains, black bars indicate predicted mouse transmembrane domains. Blue bars indicate predicted N-glycosylation sites

(consensus sequence N-X-T) in the first intravesicular loop of Lymnaea and mouse Syp. The sequence identity of Lymnaea and mouse synaptophysin homologues is ~40%. As is found in other species, synaptophysin transmembrane regions are highly conserved, and intravesicular loops show higher rates of amino acid substitutions269. The repetitive C-terminal domain is absent in

Lymnaea Syp, however the absence of the repetitive C-terminal domain is also seen in the Aplysia

Syp-like protein (accession no. XM013082245).

184

Figure 4.5 - Sequences of Lymnaea synaptophysin splice variants

185

Figure 4.5: Sequences of Lymnaea synaptophysin splice variants

(A). Protein sequence alignment of Lymnaea Syp splice variants. Grey bars indicate predicted transmembrane domains. Based on the available EST sequence information (accession no.

ES572211), I cloned Lymnaea Syp and identified 2 splice variants, which results in a truncation of 8 C-terminal residues that I designate Lymnaea synaptophysin-L (long) and Lymnaea synaptophysin-S (short). Syp mRNA microinjection experiments reported in text were performed with the Syp-L isoform. (B). Region of mRNA coding sequence variability between Lymnaea Syp-

L and Syp-S. Residues in red depict terminal stop codons. Sequence specificity of the variable region was not sufficient to permit qPCR analysis with distinct primer sets, so qPCR expression profiling was performed with primers that did not distinguish between the two Syp isoforms.

186

Figure 4.6 - Sequences of Lymnaea synaptotagmin I C2B-α and C2B-β splice variants

(A). Protein sequence alignment of Lymnaea synaptotagmin I (Syt) splice variants, C2B-α and

C2B-β (Syt-α and Syt-β). Grey bar indicates predicted transmembrane domains. Region of amino acid sequence variability is shaded grey. (B). Region of mRNA encoding sequence variability between Lymnaea Syt I splice variants is shaded grey. The alternate exon encodes 36 nucleotide changes over a region of 112 bases in the C2B domain.

187

Figure 4.7 - Prediction of Lymnaea synaptotagmin I C2B-α transmembrane domain

188

Figure 4.7: Prediction of Lymnaea synaptotagmin I C2B-α transmembrane domain

(A-B). Amino acid residue hydrophobicity plots predict the membrane topology and location of

transmembrane helices in Lymnaea Syt I C2B-α (A) and mouse Syt I (B; accession no. D37792).

Blue indicates predicted cytoplasmic residues, red indicates predicted transmembrane residues, pink indicates predicted intravesicular residues (TMHMM 2.0; CBS). (C). Protein sequence alignment of Lymnaea (top) and mouse (bottom) Syt I (Clustal Omega; EMBL-EBI). Grey bar indicates the predicted Lymnaea transmembrane domain, black bar indicates the predicted mouse transmembrane domain. The sequence identity of Lymnaea and mouse synaptotagmin I homologues is ~60%.

189

Figure 4.8 - Presynaptic synaptophysin expression is regulated by neurotrophic factors and

postsynaptic target cell identity

190

Figure 4.8: Presynaptic synaptophysin expression is regulated by neurotrophic factors and postsynaptic target cell identity

(A). Cytoplasm isolation approach for single-cell qPCR expression profiling of cultured neurons.

Scale bar, 20 μm. N=7-10 cells, 3 qPCR triplicate replicates. (B-C). Expression of FMRFa and

SVP transcripts in unpaired VD4 neurons, VF-paired VD4, and RPeD1-paired VD4 cultured in

CM (B), or DM (C). 4-FMRF, FMRFamide tetrapeptide transcript; 7-FMRF, FMRFamide heptapeptide transcript; SYP, synaptophysin; SYT α, synaptotagmin I C2B-α; SYT β, synaptotagmin I C2B-β; SYB, synaptobrevin; SYX, syntaxin. (D-E). Expression of FMRFa and

G protein transcripts in VD4-paired VF neurons (D), or VD4-paired RPeD1 neurons (E), cultured in CM or DM. Green outlines indicate culture conditions that facilitate peptidergic transmission.

Red outlines indicate culture conditions that inhibit peptidergic transmission. Dashed lines indicate transcripts were below the detection threshold. Expression was normalized to the reference gene

18s rRNA. Error bars, SEM.

191

4.4.4 Synaptophysin inhibits neuropeptide release

I next sought to determine the functional influence of regulated Syp expression on peptidergic synaptic transmission. To this end, VD4-VF neurons were cultured in CM (where VD4 exhibits low Syp expression, see Fig. 4.8B), and VD4 was microinjected with synthetic Syp mRNA to achieve overexpression. As controls I also microinjected H2O vehicle, Syt-α or Syt-β mRNA. I first measured the amplitudes of ACh mediated fast PSPs to determine whether these SVPs might differentially influence small synaptic vesicle (SSV) or LDCV fusion. Syp and Syt-β did not alter cholinergic transmission (Fig. 4.9A-B; n≥8; P≥0.400, one-way ANOVA; Table 4.12), but Syt-α decreased cholinergic transmission (n=8; P=0.009), consistent with previous observations in

Aplysia270. I then measured FMRFa mediated excitation in the presence of AChR antagonists, and

used the number of action potentials induced in VF as a measure of neuropeptide release from

VD4. Overexpression of Syp decreased peptidergic transmission (Fig. 4.9C-D; n=7; P=0.012), but

Syt-α and Syt-β had no effect (n≥8; P≥0.477). To further confirm this Syp-mediated inhibition of neuropeptide release, I also evaluated the effect of Syp overexpression on inhibitory FMRFa transmission between VD4-RPeD1 pairs in DM (where VD4 exhibits low Syp expression, see Fig.

4.8C). Syp mRNA microinjection similarly reduced the number of VD4-RPeD1 pairs exhibiting peptidergic transmission (Fig. 4.9E-F; n=2/9 peptidergic; P=0.017, Chi-squared test; Table 4.13).

These data demonstrate that Syp acts as an inhibitory regulator of LDCV, but not SSV release.

192

Figure 4.9 - Synaptophysin inhibits neuropeptide release

193

Figure 4.9: Synaptophysin inhibits neuropeptide release

(A). VD4-VF pairs were cultured in CM, and VD4 was microinjected with H2O (vehicle control),

Syp, Syt-α or Syt-β mRNA. Using intracellular current clamp recordings, the mean amplitude of

5 consecutive PSPs in VF was measured to assess changes in cholinergic synaptic transmission.

Insert shows representative PSP traces. (B). Summary data, mean PSP amplitudes of VD4-VF synapses, Syt-α reduces SSV release. N≥8. Error bars, SEM. *, P<0.05; **, P<0.01 (one-way

ANOVA). (C). The mean number of action potentials induced in VF was measured to assess changes in peptidergic synaptic transmission (isolated with AChR antagonists: 5 μM MLA; 10 μM

TC; 20 μM TEA), as in (A). (D). Summary data, mean number of action potentials in VF, Syp reduces LDCV release. N≥7. Error bars, SEM. *, P<0.05; **, P<0.01 (one-way ANOVA). (E).

Microinjection of Syp mRNA into VD4 inhibits peptidergic synaptic transmission of VD4-RPeD1 synapses in DM. (F). Summary data, incidence of VD4-RPeD1 synapses exhibiting peptidergic transmission. N≥9. *, P<0.05 (Chi-squared test; CM and DM control data is repeated from Fig.

4.1). (G). Live cell imaging of VD4-RPeD1, GFP mRNA was microinjected into VD4, GFP expression is specific to the microinjected neuron. N=5. Scale bar, 20 μm.

194

4.4.5 Trans-synaptic retrograde signaling by arachidonic acid metabolites regulates

synaptophysin expression and presynaptic peptide transmitter specificity

Since VD4 Syp expression patterns seemed to be influenced by cell-cell signaling specific to a signal derived from VF (see Fig. 4.8), I next asked whether this might involve retrograde synaptic

transmission elicited by the postsynaptic actions of FMRFa. Trans-synaptic retrograde signaling by the activity-dependent generation of membrane-permeable second messengers, such as AA or nitric oxide (NO), are known to influence synaptic function and have been previously linked to the actions of FMRFa neuropeptides171,271. To isolate the effect of VF-derived molecular signals from those of NTFs, VD4-VF neurons were cultured in DM, and I used a variety of inhibitors to screen for potential pathways with trans-synaptic influence (Fig. 4.10A; n=6-8 neurons, 3 qPCR triplicate replicates; Table 4.14). When VD4-VF neurons were cultured in DM+100 ng/mL PTX (Gαi/o

inhibitor), Syp expression was reduced to the low levels characteristic of unpaired or RPeD1-

paired VD4 in DM (see Fig. 4.8C), indicating that presynaptic Syp expression is regulated by the

postsynaptic FMRFa signaling cascades of VF neurons (see Fig. 4.1H). Syp expression was

similarly reduced in VD4-VF pairs cultured in DM+10 μM gallein (Gβγ inhibitor) or DM+10 μM

ETYA (phospholipase A2 [PLA2] and lipoxygenase [LOX] inhibitor), indicating that the target

cell-dependent transcriptional regulation of Syp involves the AA synthesis pathway. Gene

expression changes were specific to the regulation of Syp, as FMRFa and other SVP transcript

levels remained consistent amongst samples. Syp expression was unchanged in VD4-VF pairs

cultured in DM+1 μM AA, likely indicating a ceiling effect for AA signaling, or DM+10 μM L-

NNA (NG-Nitro-L-arginine; NO synthase [NOS] inhibitor), indicating that NO signaling is not

involved.

195

I then cultured VD4-VF pairs in DM+10 μM gallein or ETYA, and after ~1h washout, recorded

from VD4-VF in the presence of AChR antagonists to determine whether reduced Syp expression

would result in the formation of neuropeptide release competent synapses. In line with my

hypothesis that Syp inhibits LDCV release, downregulation of Syp induced by gallein (Fig. 4.10B-

C; n=8/12 peptidergic; P=0.003, Chi-squared test; Table 4.15) or ETYA (n=9/12 peptidergic;

P=0.001) facilitated peptidergic synaptic transmission, relative to vehicle control (DM+0.1%

DMSO; n=1/12 peptidergic). These observations implicate trans-synaptic retrograde signaling by

AA in the regulation of presynaptic Syp expression and the establishment of presynaptic peptide

transmitter specificity during synaptogenesis.

196

Figure 4.10 - Peptide transmitter specificity is induced by arachidonic acid metabolite

retrograde signaling

197

Figure 4.10: Peptide transmitter specificity is induced by arachidonic acid metabolite retrograde

signaling

(A). Expression of FMRFa and SVP transcripts in VF-paired VD4 neurons cultured in DM,

DM+100 ng/mL PTX (Gαi/o inhibitor), DM+10 μM gallein (Gβγ inhibitor), DM+10 μM ETYA

(PLA2+LOX inhibitor), DM+1 μM AA, or DM+10 μM L-NNA (NOS inhibitor). N=6-8 cells, 3 qPCR triplicate replicates. Green outlines indicate culture conditions predicted to facilitate

peptidergic transmission, red outlines indicate culture conditions predicted to inhibit peptidergic

transmission (according to Syp expression levels). Expression was normalized to the reference

gene 18s rRNA. Error bars, SEM. (B). Intracellular current clamp recordings, VD4-VF pairs in

DM do not exhibit peptidergic transmission (left). VD4-VF pairs in DM where arachidonic acid

metabolite synthesis pathways have been inhibited (gallein, middle; ETYA, right) exhibit

peptidergic transmission (isolated with AChR antagonist cocktail: 5 μM MLA; 10 μM TC; 20 μM

TEA). (C). Summary data, incidence of synapses exhibiting peptidergic transmission. ** P<0.01

(Chi-squared test).

198

4.4.6 Cell type-specific profiles of synaptophysin expression

I have so far characterized a role for Syp in the inhibitory regulation of neuropeptide release, and next asked whether cell type-specific patterns of Syp expression might be important to the regulation of neuropeptide signaling in the CNS. I performed α-FMRFa immunohistochemistry

(IHC) and Syp in situ hybridization (ISH) to determine whether there might be regional differences in Syp expression along the lines of FMRFa expression (Fig. 4.11A; n=4). A number of neurons in FMRFamidergic clusters (e.g. VF group neurons) showed little to no Syp hybridization, and this variability was not evident with non-peptidergic clusters or the other probes evaluated, including

28s rRNA, Syt-α, or Syt-β (Fig. 4.11B; n=4-8; see also Fig. 4.12). To further evaluate these apparently cell type-specific Syp expression profiles, I performed qPCR expression profiling of identified peptidergic and non-peptidergic neurons (Fig. 4.11C-D; n=6 neurons each, 3 qPCR triplicate replicates; Tables 4.16-4.17). In line with my ISH data, peptidergic VF group neurons exhibited the lowest level of Syp expression, suggesting that cell type-specific transcriptional regulation of Syp, perhaps by the local actions of AA, may allow neuropeptides to act as primary mediators of synaptic transmission in the absence of LDCV inhibition. Peptidergic VD4 and neuroendocrine CDCs (cerebral caudodorsal cells, egg laying hormone [ELH]) exhibited the high levels of Syp expression characteristic of LCDV inhibition (see Figs. 4.8-4.10). This suggests that

Syp regulation of LDCVs may ensure against ectopic neuropeptide release, and also that LDCV inhibition might be alleviated by Syp modifications in response to appropriate stimuli (e.g. synapse-specific AA signaling, or bursting activity). Non-peptidergic RPeD1 and PeA (Pedal A group) neurons also exhibited high levels of Syp expression. As these non-peptidergic cells express detectable levels of FMRFa transcripts, Syp might act to prevent any peptides that are expressed

199

from being effectively released in order to avoid the generation of confounding synaptic signals.

On the one hand, these results suggest that Syp expression levels influence the neuropeptide secretory capacity of neurons, and on the other, suggest that regulated neuropeptide release might occur irrespective of the levels of Syp transcriptional activation by a secondary active process.

200

Figure 4.11 - Cell type-specific profiles of synaptophysin expression

201

Figure 4.11: Cell type-specific profiles of synaptophysin expression

(A). IHC characterization of FMRFa immunoreactive neurons in the Lymnaea CNS (i; n=4), arrows depict the locations of FMRFa immunoreactive VF neurons (ii), and FMRFa immunonegative RPeD1 (iii) and PeA neurons (iv). (B). ISH of Syp in non-peptidergic PeA cells in the left and right pedal ganglia (i; see * in Ai) or peptidergic VF group cells in the visceral ganglia (ii; see ** in Ai). Arrows, peptidergic cells exhibiting weak Syp hybridization.

Hybridization signals did not vary between non-peptidergic and peptidergic neurons for 28s rRNA

(iii, iv), Syt-α (v, vi), or Syt-β (vii, viii). Scale bars, 100 μm. (C). Schematic of the Lymnaea CNS, depicting the locations of peptidergic (green) and non-peptidergic (grey) neurons used for in vivo single cell gene expression profiling. Dorsal surface of the ganglia: L/RCe, left and right cerebral;

L/RPe, left and right pedal; L/RPl, left and right pleural; L/RP, left and right parietal; V, visceral.

(D). In vivo expression profiles of FMRFa and SVP transcripts in peptidergic (VD4, VF, CDC; green outline) or non-peptidergic (RPeD1, PeA; red outline) neurons. N=6 cells, 3 qPCR triplicate replicates. Expression was normalized to the reference gene 18s rRNA. Error bars, SEM.

202

Figure 4.12 - In situ hybridization in the Lymnaea CNS

Negative control reactions for ISH. Hybridization of Lymnaea CNS with 28s rRNA (i, ii), Syp (iii, iv), Syt-α (v, vi) and Syt-β (vii, viii) antisense probes (top, positive) or sense probes (bottom, negative). No hybridization signals were observed with sense probes, indicating specificity of the reaction

203

4.4.7 Presynaptic peptide transmitter specificity is established notwithstanding competing

molecular signals from distinct postsynaptic targets

My results from in vitro preparations have so far demonstrated a Syp-mediated, NTF- and target

cell-dependent regulation of neuropeptide release from VD4. However, VD4 neurons in vivo

exhibit the high levels of Syp expression characteristic of peptide non-release. I next asked whether

this form of presynaptic transmitter specificity might be established in the presence of competing

molecular signals from different postsynaptic targets. I first took advantage of the bipolar

morphology of VD4 to culture triple axon pairs with both VF and RPeD1 postsynaptic targets in

CM, and performed FMRFa ICC. The VD4-RPeD1 synaptic terminals, but not the VD4-VF

synaptic terminals, exhibited the ‘hyper-innervation’ phenotype characteristic of peptide non- releasing synapses, where intensity of the FMRFa immunolabel was significantly higher (Fig.

4.13A-B; n=8; P=0.004, independent t-test; Table 4.18). Simultaneous intracellular recordings from triple soma pairs cultured in CM demonstrated that VD4 formed a mixed cholinergic- peptidergic synapse with VF, where most VD4 neurons released both ACh and FMRFa (Fig.

4.13C-D; n=7/8 peptidergic; Table 4.19), whereas VD4 formed a purely cholinergic synapse with

RPeD1, where all VD4 neurons released ACh, but few released FMRFa (n=2/8 peptidergic;

P=0.012, Chi-squared test). These results suggest that FMRFa LDCVs at VD4-VF synaptic terminals are release competent, whereas the ones at VD4-RPeD1 terminals are not.

I next asked whether presynaptic peptide transmitter specificity might also occur in vivo despite the elevated Syp expression that would seemingly preclude neuropeptide release. I performed simultaneous 3 cell recordings from the isolated intact CNS (Fig. 4.14A-C), and monitored the

204

postsynaptic responses of VF and RPeD1 to VD4 stimulation. Whereas cholinergic transmission

occurred at both VF and RPeD1 synapses, peptidergic transmission occurred at VF synapses (Fig.

4.14D-E; n=10/11 peptidergic; Table 4.19) but not at RPeD1 synapses (n=3/11 peptidergic;

P=0.002, Chi-squared test). These observations suggest that FMRFa neuropeptide release is

facilitated at endogenous VD4 presynaptic terminals innervating VF, but inhibited at presynaptic

terminals innervating RPeD1. Taken together, these data demonstrate that the synapse-specific use

of peptide neurotransmitters occurs in vivo notwithstanding elevated Syp expression and

competing molecular signals from postsynaptic targets, and is appropriately recapitulated by cell-

cell molecular signaling interactions in vitro. These data thus establish a requirement for synapse-

specific populations of release competent or incompetent LDCVs in regulated peptidergic transmission.

The above observations support a novel model for the regulation of synaptic neuropeptide release in which (i) transcriptional regulation of Syp is functionally significant when the synaptogenic molecular cues a neuron receives from the environment and postsynaptic target(s) are homogeneous, but (ii) when faced with heterogeneous synaptogenic signals, the Syp transcriptional program is superseded by a secondary active process which produces presynaptic specificity, wherein local variations in NTFs and target cell-specific molecular signals adjust the neuropeptide release competency of individual synapses to meet the functional requirements for synaptic transmission within neuronal networks (Figs. 4.15-4.16).

205

Figure 4.13 - Peptide transmitter specificity occurs with competing postsynaptic targets

206

Figure 4.13: Peptide transmitter specificity occurs with competing postsynaptic targets

(A). ICC labeling of FMRFa neuropeptides in triple-axon paired RPeD1-VD4-VF. Arrowheads, putative peptide releasing synaptic sites. Arrows, putative peptide non-releasing synaptic sites.

Scale bar, 50 μm. (B). Summary data, fluorescence intensity is increased at synapses that do not exhibit peptidergic transmission. Error bars, SEM. **, P<0.01 (independent t-test). (C).

Intracellular current clamp recordings from neurons cultured in a triple-soma configuration in CM

(insert; scale bar, 20 μm). VD4 forms a purely cholinergic synapse with RPeD1 and a mixed cholinergic-peptidergic synapse with VF. Inserts show ACh-mediated transmission (PSPs, -100 mV holding potential), inhibited by AChR antagonists (5 μM MLA, 10 μM TC, 20 μM TEA). (D).

Summary data, incidence of synapses exhibiting peptidergic transmission. *, P<0.05 (Chi-squared test).

207

Figure 4.14 - Peptide transmitter specificity occurs in vivo despite elevated synaptophysin expression

208

Figure 4.14: Peptide transmitter specificity occurs in vivo despite elevated synaptophysin expression

(A-C). Lucifer yellow dye injection of RPeD1 (A), VD4 (B), and VF neurons (C), to illustrate the isolated intact preparation and morphology of in situ projections. Scale bar, 1 mm. (D).

Intracellular current clamp recordings from endogenous VD4 synapses in an isolated intact CNS preparation. In vivo, VD4 expresses the high levels of Syp characteristic of LDCV inhibition (see

Fig. 4.11D), but VD4 forms a purely cholinergic synapse with RPeD1 and a mixed cholinergic-

peptidergic synapse with VF in vivo, indicating that synapse-specific populations of release

competent or incompetent LDCVs underlies this form of peptidergic presynaptic specificity. (E).

Summary data, incidence of synapses exhibiting peptidergic transmission. **, P<0.01 (Chi-

squared test).

209

Figure 4.15 - A model for neurotrophic factor and cell-cell signaling in synaptophysin-

dependent presynaptic peptide transmitter specificity

210

Figure 4.15: A model for neurotrophic factor and cell-cell signaling in synaptophysin-dependent

presynaptic peptide transmitter specificity

(A). NTF signaling via RTKs stimulates Syp transcription, elevated Syp expression inhibits

neuropeptide release. (B). In the absence of NTF/RTK activation, reduced Syp expression

facilitates neuropeptide release (e.g. VD4-RPeD1 paired neurons, see Figs. 4.1-4.9). (C).

Postsynaptic activation of Gαi/o coupled FMRFa GPCRs, leading to excitation, stimulates

production of AA via PLA2. AA acts trans-synaptically to counteract the influence of NTF/RTK

activation and inhibits Syp transcription. Reduced Syp expression facilitates neuropeptide release.

(D). In the absence of NTF/RTK signaling, retrograde AA signaling stimulates Syp transcription, elevated Syp expression inhibits neuropeptide release (e.g. VD4-VF paired neurons, see Figs. 4.1-

4.10).

211

Figure 4.16 - A model for posttranslational modification of synaptophysin in synapse-

specific inhibition of neuropeptide release

212

Figure 4.16: A model for posttranslational modification of synaptophysin in synapse-specific

inhibition of neuropeptide release

(A). With competing synaptogenic molecular signals (e.g. VD4 innervating VF and RPeD1, see

Fig. 4.13-4.14), synapse-specific release of neuropeptides could be established by the differential

trafficking of Syp-poor LDCVs to neuropeptide release-competent terminals and Syp-rich LDCVs to neuropeptide release-incompetent terminals, N-glycosylation may act as a selectivity filter for the differential sorting of Syp to LDCVs272. (B). Synapse-specific release of neuropeptides could be established by local posttranslational modifications of Syp that either promote the dissociation of Syb for release competent LDCVs (AA-PKC?)172, or inhibit the dissociation of Syb for release

incompetent LDCVs (RTK-GSK3β?)273. Activity-dependent modification of Syp might also

switch release incompetent LDCVs to release competent LDCVs to promote context-dependent

neuropeptide release (e.g. CDC neurons express high levels of Syp (see Fig. 4.11), and only release

ELH in response to prolonged bursting activity274).

213

4.5 Discussion

While all of a neuron’s presynaptic terminals are linked to the transcriptional activity of a single

nucleus, the composition and function of each synapse can be regulated individually. My results

demonstrate a novel role for Syp in establishing synapse-specific use of co-transmitters via the inhibitory regulation of neuropeptide release. This mechanism of presynaptic transmitter specificity is dynamically regulated by both extrinsic NTFs and target cell-specific retrograde signaling via AA, indicating that it is a previously undescribed form of context-dependent synaptic specificity and plasticity that tunes the release competency of individual presynaptic terminals.

Despite its prevalent expression and localization at synaptic terminals throughout the CNS, Syp is not essential for SV release16, and the functional significance of the Syp family of SVPs has remained obscure. On one hand, my data from 1:1 neuron pairs demonstrate that transcriptional regulation of Syp by NTFs and AA can fully account for the formation of neuropeptide release- competent or -incompetent synapses (Figs. 4.1-4.10). On the other hand, my data from triple soma and intact synapses indicate that a secondary active process occurs to selectively facilitate or inhibit neuropeptide release locally in the presence of competing synaptogenic cues (Figs. 4.11-4.14).

This is likely established by (i) the differential trafficking of release competent (i.e. Syp-) or release

incompetent (i.e. Syp+) LDCVs, or (ii) synapse-delimited posttranslational modification (PTM)

of Syp to alter the release competency of Syp+ LDCVs. There is some evidence to implicate both

of these processes in regulating the variability of peptidergic presynaptic terminals.

214

4.5.1 Differential sorting and trafficking of large dense-core vesicles

There have been conflicting reports suggesting that Syp is absent from LDCVs21, present at lower

concentrations than SSVs (~10%)275, or present in varying amounts276. In mouse neurons, N-

glycosylation, a PTM occurring in the endoplasmic reticulum that affects protein sorting and

trafficking, has been shown to be essential for targeting Syp, but not Syt I, out of the cell body and

to the synapse272. This observation suggests that N-glycosylation acts as a selectivity filter to

govern the amount of Syp incorporated during LDCV biogenesis, and thereby regulates the release

competency of LDCVs. Downstream of biogenic sorting, the synapse-specific targeting of cargo,

including LDCVs, is a second highly regulated process that differentiates the composition and

function of individual presynaptic terminals during synaptogenesis and synaptic plasticity. This is

mediated by molecular motors that have the capacity to recognize appropriate synaptic cargo and

direct transport to relevant subcellular compartments, such as synapses that have been molecularly

tagged by activity66,125,277. My data therefore suggest that the controversial, highly variable

presence of Syp on LDCVs might be tuned by the influence of NTF and AA signaling on N- glycosylation. Taken together, this offers a putative mechanism through which the synapse- specific use of neuropeptides and plasticity of peptidergic synaptic transmission might be achieved.

4.5.2 Context-dependent posttranslational modifications of synaptophysin

Syp monomers assemble into hexameric channels that orient Syb dimers to promote the SNARE

interactions necessary for SV fusion278. Whereas formation of the Syb-Syp complex is dispensable

for SV exocytosis, the release of Syb from Syp is prerequisite for SNARE zippering18. The

215

observation that this dissociation can be inhibited by glycogen synthase kinase 3β (GSK3β)-

dependent phosphorylation in rat hippocampal neurons suggests that cytoplasmic PTM of Syp can

also act to regulate the release competency of SVs273. In the mammalian brain, formation of the

Syb-Syp complex, due to Syp PTM, has been reported to be upregulated during development279, and absent in neuroendocrine cells280, revealing that Syp-regulated SV release might be selectively

governed by a wide variety of extrinsic and intrinsic molecular cues that are yet to be determined.

These observations of modulatory Syb-Syp affinity, and the resulting inhibition or facilitation of

SV release, can be extended to my results on the context-dependent regulation of LDCV release.

For instance, it has been suggested that Syb-Syp interactions may act to establish a reserve pool of

SVs that can be recruited by synaptic activity279. Applied to LDCVs, this adjustable reserve

function would have interesting applications to our understanding of the priming and typically

burst-dependent release dynamics of neuropeptides or neurohormones, as well as the occasional

instances observed in invertebrate neurons where a single action potential is sufficient to elicit

neuropeptide release138. It should, however, also be noted that the presynaptic terminals of co- transmitting neurons usually do not contain large numbers of LDCVs, and transiting LDCVs have been shown to be rapidly recruited by activity in Drosophila motoneurons66. This suggests that

peptidergic synaptic transmission is effectively maintained in the absence of a substantial LDCV

reserve pool, and perhaps underlines the significance of a mechanism for differential sorting and

trafficking.

216

4.5.3 Extracellular and cell-cell signaling in peptidergic synaptic specificity

Here, I found that presynaptic transmitter specificity of VD4 was established by synapse-specific

induction of AA, where one postsynaptic target was excited by FMRFa (VF, via Gαi/o), and another

2+ was inhibited (RPeD1, via non-Gαi/o). Ca -dependent activation of PLA2 is required for the release of AA from membrane phospholipids, and is therefore conditional on an excitatory postsynaptic response resulting in Ca2+ influx, although I do not exclude the possible involvement

of Ca2+ release from internal stores via second messenger signal transduction. The effects of AA

retrograde signaling on synaptic plasticity have been found to be mediated directly via inhibition

of voltage-gated ion channels or the activation of protein kinase C (PKC)4,172, although it is

possible that other molecular effectors of AA signaling are yet to be identified. Competing kinase

signaling cascades initiated by AA (e.g. PKC) or NTF (e.g. mitogen-activated protein kinase

[MAPK] cascades), which activate or inhibit effector proteins that act locally at the synapse or

translocate to the nucleus to regulate gene expression, are likely how regulated Syp expression and

synapse-specific function is achieved. It is of interest to note that whereas NTF signaling is well

known to influence gene expression71, to the best of my knowledge, transcriptional regulation has

not previously been implicated in the effects of AA on synaptic plasticity. I found AA signaling to

promote Syp expression in the absence of NTF, and inhibit it in the presence of NTF, which

suggests that AA likely influences the activity of transcription factors, perhaps by inducing a

phosphorylation signal for activation via PKC, or masking an activation signal provided by NTFs.

As synapse-specific variations in the peptidergic characteristics of invertebrate co-transmitting neurons have been described for a variety of neuropeptides (e.g. FMRFa137,259, small cardioactive

217

peptides138, and proctolin139), the Syp-dependent mechanism of inhibitory regulation of

neuropeptide release identified here may be a universal, adaptable mechanism that governs the

peptidergic characteristics of presynaptic terminals. While a prerequisite for my proposed model

of presynaptic specificity is target-dependent differences in molecular signaling, synapse-specific

release of the neuropeptide proctolin from a Cancer presynaptic neuron has also been reported to occur with target neurons that both exhibit postsynaptic excitation139. While this is presumably

due to the actions of proctolin acting on the same GPCR, presynaptic differences in peptide release

competency could emerge as a result of (i) dissimilar profiles of postsynaptic effector proteins

such as PLA2 and LOX, (ii) differential localization of effector proteins (e.g. PKC) in the

presynaptic terminal, or (iii) local variations in NTF signaling, for instance, as a result of target- specific expression patterns of NTFs. When taken together, the above observations support a role for synapse-specific molecular signaling in the dynamic regulation of presynaptic peptidergic characteristics.

4.5.4 Functional implications of synaptophysin-regulated peptidergic transmission

FMRFa and FMRFamide related peptides (FaRPs) are encoded by a highly conserved gene family that is found throughout the animal kingdom. In mammals, there are currently five identified

FaRP-encoding genes, whose protein products (e.g. prolactin-releasing peptide, kisspeptin) play important roles in the synaptic and physiological changes underlying reproductive behaviour167.

Neuropeptide Y (NPY) is the most widely expressed neuropeptide in the mammalian brain, and its aberrant expression and release has been implicated in the emergence of synaptic plasticity deficits underlying the etiology of neuropsychiatric conditions such as depression281. NPY shares

218

a homologous C-terminal structure and high degree of GPCR sequence conservation with FaRP

systems, suggesting that they are both evolutionarily and functionally related282. As such, the

present findings regarding the regulation of peptidergic synaptic transmission by Syp and the

influence of invertebrate FMRFa neuropeptides on synapse formation and synaptic specificity are

likely to be translatable to the roles of mammalian FaRPs and NPY in the brain.

While members of the Syp family of SVPs are known to be essential regulators of synaptic plasticity19, the molecular mechanism underlying this effect has not been determined. This study

suggests that the selective inhibitory regulation of peptide neurotransmitters may be responsible

for Syp-dependent synaptic plasticity. The present data indicate that NTF regulation of Syp

underlies positive or negative changes in peptidergic transmission, according to the metabotropic

response of the postsynaptic target. This finding identifies a new component of NTF-induced

synaptic plasticity, in mediating the functional differentiation and plastic remodeling of individual

presynaptic terminals in a target cell-specific manner. Considering that neuropeptides are important modulatory signals for the synaptic plasticity events underlying learning and memory,

it is intriguing that (i) NTFs are implicated in the synaptic plasticity deficits of neuropsychiatric conditions such as depression283 and schizophrenia284, and (ii) Syp is one of the genetic mutations

implicated in X-linked intellectual disability285. Further characterization of this Syp-dependent

mechanism for regulated neuropeptide release, as it pertains to synaptic maturation, specificity and

plasticity, is likely to offer new insights into the etiology of these neurodevelopmental and

neuropsychiatric disorders.

219

Table 4.1 - Lymnaea cloning primers

Target Accession 5’ Primer Sequence 3’ Primer Sequence Number SYP ES572211 GATGGGCCCATGGAGGC GATGAATTCTCAAGCTGTGT CGCCGGACAG GGGACTCTAC SYT α / β AF484090 GATCTCGAGATGCCTGCC GATGGATCCTTAGCTCTTCTC CTG TG eGFP - GATGAATTCATGGTGAG GATGGATCCCTAGCTACTAG CAAGGGCGAGG CTAGTCGAG

Table 4.2 - Lymnaea RT-PCR gene specific primers

Target Accession 5’ Primer Sequence 3’ Primer Sequence Number 18s rRNA Z73984 CTGGTTGATCCTGCCAG CTTCCGCAGGTTCACCTAC TAG FMRF 4 / 7 M87479 GATCTCGAGATGAAAAC GATGGATCCTCATCGAGA GTGGAGTCACGTG CTGTTCGGCCCC SYP ES572211 GATGGGCCCATGGAGGC GATGAATTCTCAAGCTGTG CGCCGGACAG TGGGACTCTAC SYT α / β AF484090 ATGCCTGCCCTGGGCGC TTAGCTCTTCTCTGGCACT TC SYB AF484091 ATGGCTGCTTCCCAAAA CTAAGTTGTCTTAGGTGAT CCC GG SYX AF484088 ATGACGAAGGATAGATT GTCAACCAAAAGTCCCGC AGCTG C G α q Z23106 GATGTCGACATGGCCTG GATTCTAGATCACACCAA TTGTATTCCGGAT GTTGTACTCTTTC G α i Z15095 GATCTCGAGATGGGGTG GATGGATCCTTAGAAAAG TGTAACCAGCC GCCGCAATCCTT G α o Z15094 GATCTCGAGATGGGGTG GATGGATCCTCAGTAGAG TACTCTGAGCG CCCACAACCC G α s Z15096 GATCTCGAGATGGGGTG GATTCTAGATCACAGAAG CTTTCGACAAACC CTCGTATTGTCG G α a Z47551 GATCTCGAGATGAGCAA GATGAATTCTTAATAATTC AGGTGGGCGGG TTCGACCGTTTCAC G β Z23105 GATCTCGAGATGAGTAA GATGGATCCTCAGTTCCAG CGATTTAGAAGCCTT ATCTTCAAAAAGC

220

Table 4.3 - Lymnaea qPCR gene specific primers

Target Accession 5’ Primer Sequence 3’ Primer Sequence Efficiency Number (%) 18s rRNA Z73984 CACGGGGAGGTAGT GCCCTCCAATGGGTC 103.40 GACG CTC 4-FMRF M37629 TGTGCAGTATGACCA CGGCCAATTCGATAG AACGC AACTG 111.42 7-FMRF U03137 GATATGTGCAGTAAC ATCCCCGTCTTCTTCT TCTGTG GGG 101.79 SYP ES572211 GCGCAACACCAGGA TCAGACACCCCAGAT TTGTC GCCC 104.94 SYT α - ATTGCTCTTGTACAG TCCAAAAGACTCGTT GGCAC GAAATATG 96.45 SYT β AF484090 AAATCTCTCTGATGC CGTAAAAGATTCATT TCAATGG ATAGTAAGG 90.65 SYB AF484091 CGTCTGCAACAGACC CTGCTCTATCATCCAG CAAGC TTGAG 105.36 SYX AF484088 AAGTCAGGGTGAAA CCGCCGCGCTTTACTT TGATAGAC TGG 108.01 G α q Z23106 TATATTCATGGCGAT TGTCTCGTAATCAATT GCATGC TGACGG 91.36 G α i Z15095 ACGCCCGGCAGTTTT ACACTCTTGTACACC TCGC AACATC 95.86 G α o Z15094 GATGTTATAGCCAGG TACTCATTGGCACGA ATGGAG CCTAAG 94.16 G α s Z15096 TCGACAAACCAGAG CAACAACCTATGTGT ACGACG TCCTCTG 87.48 G α a Z47551 GAGAGGCGAGAGAA CTCGTTGGGCGGGAA AGAGG CTG 85.39 G β Z23105 GATGGCTACACAAC GTCCAAGGCATCGGA AAATAAGG GTAGC 94.44

221

Table 4.4 - Lymnaea in situ hybridization probe sequences

Target Riboprobe Sequence 28s rRNA antisense Refer to Bauman & Bentvelzen P, 1988286 and van Minnen et al., 28s rRNA sense 1997261. GAATCTGTTCAAATGGAACCTCAAATCCAAAAGACTCGT SYT α antisense TGAAATATGGATTGAGGGTGTTTTTCTTAATAGTCGTTTT CTTTTTCTTGAGACGCTTTGTGCCCTGTACAAGAGCAATT AATTGCTCTTGTACAGGGCACAAAGCGTCTCAAGAAAAA GAAAACGACTATTAAGAAAAACACCCTCAATCCATATTT SYT α sense CAACGAGTCTTTTGGATTTGAGGTTCCATTTGAACAGAT TC TTGTTCGAATGGAACTTCAAACGTAAAAGATTCATTATA SYT β antisense GTAAGGATTGAGAGTGCATTTCTTGATGGTTGTCTTCTTC TTTTTCACTCGTTTTCCATTGAGCATCAGAGAGATTT AAATCTCTCTGATGCTCAATGGAAAACGAGTGAAAAAG SYT β sense AAGAAGACAACCATCAAGAAATGCACTCTCAATCCTTAC TATAATGAATCTTTTACGTTTGAAGTTCCATTCGAACAA SYP antisense Full length cDNA sense/antisense sequence, see accession number SYP sense ES572211.

Table 4.5 - Peptidergic characteristics of VD4-VF and VD4-RPeD1 synapses

CM FMRF Releasing χ2 P 1 N VD4-VF 8/8 (100%) 8 9.164 .012 VD4-RPeD1 3/10 (30%) 10 DM FMRF Releasing χ2 P N VD4-VF 2/10 (20%) 10 6.600 .010 VD4-RPeD1 9/12 (75%) 12

1. Chi-squared test (2-sided)

222

Table 4.6 - Somatic FMRFa fluorescence in VD4

Culture Condition Mean Fluorescence Standard Error P 1 N (AU) (AU) CM VD4 150.48 9.57 (F=0.301; P=0.910) 7 CM VD4-RPeD1 154.58 5.51 .999 9 CM VD4-VF 153.13 7.34 1.000 12 DM VD4 147.89 5.58 1.000 11 DM VD4-RPeD1 156.64 6.88 .996 9 DM VD4-VF 160.66 13.71 .967 8

1. One-way ANOVA with Tukey’s HSD post hoc test, significance (P) relative to CM VD4

Table 4.7 - Synaptic FMRFa fluorescence

Culture Condition Mean Standard P 1 P 2 N Fluorescence Error (AU) (AU) CM VD4-RPeD1 76.28 8.62 (F=10.978; P<0.001) .972 9 CM VD4-VF 45.99 4.29 .037 .001 12 DM VD4-RPeD1 46.66 2.72 .037 - 9 DM VD4-VF 80.72 5.63 .972 .001 8

1. One-way ANOVA with Games-Howell post hoc test, significance (P) relative to CM VD4-

RPeD1

2. One-way ANOVA with Games-Howell post hoc test, significance (P) relative to DM VD4-

RPeD1

223

Table 4.8 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in VD4

Culture CM VD4 Single CM VD4 VF-paired CM VD4 RPeD1-paired Condition -ΔC 1 -ΔC -ΔC Target 2 T SEM 2 T SEM 2 T SEM 4-FMRF 7.79E-3 6.60E-3 – 1.25E-8 1.11E-8 – 3.41E-3 3.23E-3 – 9.19E-3 1.41E-8 3.59E-3 7-FMRF 2.98 2.55 – 1.02 0.96 – 2.65 2.51 – 3.50 1.09 2.80 SYP 7.24E-4 6.17E-4 – 2.31E-9 2.11E-9 – 1.17E-3 1.13E-3 – 8.48E-4 2.53E-9 1.22E-3 SYT α 1.57E-6 1.32E-6 – 1.42E-8 1.20E-8 – 3.80E-10 3.52E-10 – 1.87E-6 1.66E-8 4.11E-10 SYT β 9.57E-8 7.88E-8 – ND 2 - 2.17E-7 1.98E-7 – 1.16E-7 2.38E-7 SYB 2.49E-2 1.67E-2 – 6.79E-3 6.40E-3 – 3.53E-2 3.34E-2 – 3.72E-2 7.20E-3 3.74E-2 SYX 1.96E-7 1.69E-7 – 3.60E-7 3.41E-7 – 5.08E-8 4.50E-8 – 2.28E-7 3.80E-7 5.73E-8

1. Relative expression, normalized to 18s rRNA reference gene

2. ND indicates a signal was not detected during the qPCR reaction

224

Table 4.9 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in VD4

Culture DM VD4 Single DM VD4 VF-paired DM VD4 RPeD1-paired Condition -ΔC 1 -ΔC -ΔC Target 2 T SEM 2 T SEM 2 T SEM 4-FMRF 2.65E-5 2.50E-5 – 5.63E-5 5.29E-5 – 1.51E-5 1.39E-5 – 2.81E-5 5.99E-5 1.64E-5 7-FMRF 0.72 0.65 – 2.99 2.81 – 0.40 0.38 – 0.80 3.18 0.42 SYP 6.98E-11 6.75E-11 – 1.80E-3 1.71E-3 – 1.41E-10 1.35E-5 – 7.20E-11 1.90E-3 1.45E-10 SYT α 5.21E-8 4.83E-8 – 6.15E-6 4.83E-8 – 7.61E-11 4.96E-11 – 5.62E-8 5.62E-8 1.17E-10 SYT β 9.70E-8 8.87E-8 – 8.44E-7 8.87E-8 – 2.64E-8 2.51E-8 – 1.06E-7 1.06E-7 2.77E-8 SYB 2.92E-3 2.76E-3 – 3.32E-3 2.76E-3 – 3.13E-3 2.98E-3 – 3.09E-3 3.09E-3 3.30E-3 SYX 5.62E-7 5.31E-7 – 7.59E-7 7.17E-7 – ND 2 - 5.95E-7 8.04E-7

1. Relative expression, normalized to 18s rRNA reference gene

2. ND indicates a signal was not detected during the qPCR reaction

225

Table 4.10 - Single cell qPCR - Normalized expression of G proteins in VF

Culture CM VF VD4-paired DM VF VD4-paired Condition -ΔC 1 -ΔC 1 Target 2 T SEM 2 T SEM 4-FMRF 1.83E-3 1.46E-3 - 2.28E-3 6.44E-2 4.20E-2 - 9.87E-2 7-FMRF 8.78E-1 7.40E-1 - 1.04 3.00E-1 2.00E-1 - 4.50E-1 G α q 3.40E-7 3.00E-7 - 3.85E-7 4.42E-6 3.00E-6 - 6.52E-6 G α i 6.91E-3 6.03E-3 - 7.92E-3 2.16E-3 1.48E-3 - 3.15E-3 G α o 2.10E-4 1.85E-4 - 2.39E-4 3.63E-4 2.56E-4 - 5.15E-4 G α s 2.21E-4 1.91E-4 - 2.55E-4 3.29E-4 2.31E-4 - 4.69E-4 G α a 1.27E-11 1.07E-11 - 1.52E-11 1.88E-11 1.04E-11 - 3.40E-11 G β 2.22E-10 1.38E-10 - 3.55E-10 1.40E-6 8.80E-7 - 2.23E-6

1. Relative expression, normalized to 18s rRNA reference gene

Table 4.11 - Single cell qPCR - Normalized expression of G proteins in RPeD1

Culture CM RPeD1 VD4-paired DM RPeD1 VD4-paired Condition -ΔC 1 -ΔC Target 2 T SEM 2 T SEM 4-FMRF 3.21E-6 2.65E-6 - 3.88E-6 6.73E-7 5.18E-7 - 8.74E-7 7-FMRF 1.66E-2 1.47E-2 - 1.88E-2 4.25E-2 2.89E-2 - 6.24E-2 G α q 1.93E-6 1.85E-6 - 2.02E-6 1.93E-7 1.49E-7 - 2.51E-7 G α i 3.08E-3 2.87E-3 - 3.31E-3 2.11E-3 1.57E-3 - 2.83E-3 G α o 8.86E-3 8.37E-3 - 9.37E-3 7.88E-4 5.93E-4 - 1.05E-3 G α s 6.29E-2 5.76E-2 - 6.87E-2 1.73E-2 1.33E-2 - 2.26E-2 G α a 6.77E-9 4.84E-9 - 9.46E-9 1.16E-7 9.10E-8 - 1.48E-7 G β 5.12E-6 3.24E-6 - 8.08E-6 2.13E-6 1.53E-6 - 2.97E-6

1. Relative expression, normalized to 18s rRNA reference gene

226

Table 4.12 - ACh and FMRFa synaptic transmission characteristics of VD4-VF pairs

PSP Amplitude Standard Error P 3 N CM VD4-VF - ACh1 (mV) (mV) VD4 + H2O 11.10 1.17 (F=4.040; P=0.015) 8 VD4 + SYP mRNA 9.92 2.37 .969 10 VD4 + SYT α mRNA 5.67 0.62 .009 8 VD4 + SYT β mRNA 15.96 2.75 .400 10 VF Action Potentials Standard Error P 3 N CM VD4-VF – FMRF2 (#) (#) VD4 + H2O 4.33 0.78 (F=4.557; P=0.009) 9 VD4 + SYP mRNA 1.07 0.28 .012 7 VD4 + SYT α mRNA 6.88 1.52 .477 8 VD4 + SYT β mRNA 6.29 1.23 .545 12 VD4 Burst Duration Standard Error P 3 N CM VD4-VF – FMRF (s) (s) VD4 + H2O 4.49 0.57 (F=0.202; P=0.894) 9 VD4 + SYP mRNA 4.52 0.43 1.000 7 VD4 + SYT α mRNA 4.17 0.27 0.947 8 VD4 + SYT β mRNA 4.20 0.33 0.954 12

1. Recorded in control conditions, PSP measured at -100 mV holding potential

2. Recorded in AChR inhibitors, excitation measured at -60 mV holding potential

3. One-way ANOVA with Games-Howell post hoc test, significance (P) relative to VD4 +

H2O

227

Table 4.13 - Peptidergic characteristics of VD4-RPeD1 synapses

VD4-RPeD1 FMRF Releasing χ2 P 2 χ2 P 3 CM1 3/10 (30%) - - 4.455 .035 DM1 9/12 (75%) 4.455 .035 - - DM + SYP mRNA (VD4) 2/9 (22%) 0.148 .701 5.743 .017

1. Data is repeated here from Table 4.5

2. Chi-squared test (2-sided), significance (P) relative to CM

3. Chi-squared test (2-sided), significance (P) relative to DM

228

Table 4.14 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in VD4

Culture DM + 0.1% DMSO DM + 100 ng/mL PTX DM + 10 μM Gallein Condition + 0.1% Tocrisolve

-ΔC 1 -ΔC -ΔC Target 2 T SEM 2 T SEM 2 T SEM 4-FMRF 1.93E-3 1.51E-3 – 3.36E-6 3.15E-6 – 1.12E-4 1.06E-4 – 2.46E-3 3.58E-6 1.18E-4 7-FMRF 3.70 2.91 – 3.42 3.19 – 1.04 0.96 – 4.71 3.66 1.12 SYP 3.29E-3 2.56E-3 – 6.96E-9 5.41E-9 – 7.76E-10 5.64E-10 – 4.23E-3 8.96E-9 1.07E-9 SYT α 6.71E-6 5.24E-6 – 2.49E-6 2.34E-6 – 1.30E-7 1.16E-7 – 8.58E-6 2.64E-6 1.46E-7 SYT β 3.04E-07 2.29E-7 – 3.7E-8 3.43E-8 – 7.42E-8 6.98E-8 – 4.03E-7 3.99E-8 7.9E-8 SYB 0.20 0.15 – 4.78E-3 4.47E-3 – 1.20E-2 1.13E-2 – 0.26 5.11E-3 1.28E-2 SYX 2.08E-6 1.58E-6 – 3.85E-7 3.46E-7 – 1.79E-8 1.63E-8 – 2.74E-6 4.29E-7 1.96E-8

Culture DM + 10 μM ETYA DM + 1 μM AA DM + 10 μM L-NNA Condition

-ΔC -ΔC -ΔC Target 2 T SEM 2 T SEM 2 T SEM 4-FMRF 9.94E-6 8.62E-6 – 1.41E-3 1.19E-3 – 1.88E-5 1.56E-5 – 1.15E-5 1.68E-3 2.25E-5 7-FMRF 3.66 3.17 – 1.11 0.96 – 2.19 4.22 1.28 1.77 - 2.72 SYP 2.22E-9 1.80E-9 – 2.84E-4 2.47E-4 – 1.96 E-4 1.62E-4 – 2.75E-9 3.28E-4 2.38E-4 SYT α 3.66E-6 3.15E-6 – 5.89E-7 4.98E-7 – 1.04E-7 8.55E-8 – 4.24E-6 6.96E-7 1.26E-7 SYT β 8.33E-7 7.13E-7 – 6.21E-8 5.39E-8 – 3.77E-8 2.96E-8 – 9.75E-7 7.16E-8 4.79E-8 SYB 3.70E-2 3.16E-2 – 1.93E-2 1.67E-2 – 1.64E-2 1.37E-2 – 4.35E-2 2.22E-2 1.97E-2 SYX 2.01E-7 1.74E-7 – 9.86E-8 7.85E-8 – 7.19E-8 5.79E-8 – 2.33E-7 1.24E-7 8.91E-8

1. Relative expression, normalized to 18s rRNA reference gene

229

Table 4.15 - Peptidergic characteristics of VD4-VF synapses in DM

VD4-VF FMRF Releasing χ2 P 1 DM + 0.1% DMSO 1/12 (8%) - - DM + 10 μM Gallein 8/12 (67%) 8.711 .003 DM + 10 μM ETYA 9/12 (75%) 10.514 .001

1. Chi-squared test (2-sided), significance relative to DM + 0.1% DMSO

Table 4.16 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in peptidergic cell types

Peptidergic VD4 VF ELH Neurons -ΔC 1 -ΔC -ΔC Target 2 T SEM 2 T SEM 2 T SEM 4-FMRF 2.84E-2 1.90 E-2 – 6.77 5.62 – 4.76E-2 3.29E-2 – 4.26E-2 8.17 6.88E-2 7-FMRF 51.98 37.71 – 5.52E-6 5.04E-6 – 5.88E-8 4.38E-8 – 71.67 6.05E-6 7.89E-8 SYP 3.55E-3 2.57E-3 – 2.36E-6 1.20E-6 – 9.52E-6 7.70E-4 – 4.88E-3 4.65E-6 1.18E-3 SYT α 1.31E-7 9.06E-8 – 2.31E-6 1.32E-6 – 3.79E-8 1.82E-8 – 1.90E-7 4.03E-6 7.89E-8 SYT β 2.18E-8 1.43E-8 – 4.46E-7 3.12E-7 – 1.03E-8 7.31E-9 – 3.32E-8 6.38E-7 1.45E-8

1. Relative expression, normalized to 18s rRNA reference gene

230

Table 4.17 - Single cell qPCR - Normalized expression of synaptic vesicle proteins in non- peptidergic cell types

Non- Peptidergic RPeD1 PeA Neurons -ΔC 1 -ΔC Target 2 T SEM 2 T SEM 4-FMRF 3.27E-7 2.82E-7 - 3.80E-7 9.40E-10 8.13E-10 - 1.09E-9 7-FMRF 1.31E-2 1.01E-2 - 1.70E-2 4.00E-7 3.35E-7 - 4.78E-7 SYP 1.22E-3 1.00E-3 - 1.49E-3 3.20E-2 2.68E-2 - 3.82E-2 SYT α 3.09E-9 2.22E-9 - 4.30E-9 2.51E-7 1.89E-7 - 3.34E-7 SYT β 1.26E-9 6.55E-10 - 2.41E-9 4.61E-8 2.97E-8 - 7.15E-8

1. Relative expression, normalized to 18s rRNA reference gene

Table 4.18 - Synaptic FMRFa fluorescence in RPeD1-VD4-VF triple axon pairs in CM

1 RF-NH2 Mean Fluorescence (AU) Standard Error (AU) P N VD4-RPeD1 109.32 7.33 .004 8 VD4-VF 75.05 7.01

1. Independent samples t-test (2-sided)

Table 4.19 - Peptidergic characteristics of RPeD1-VD4-VF synapses in vitro and in vivo

RPeD1-VD4-VF in vitro FMRF Releasing (%) χ2 P 1 N VD4-RPeD1 2/8 (25%) 6.349 .012 8 VD4-VF 7/8 (88%) RPeD1-VD4-VF in situ FMRF Releasing (%) χ2 P 1 N VD4-RPeD1 3/11 (27%) 9.214 .002 8 VD4-VF 10/11 (91%)

1. Chi-squared test (2-sided)

231

Chapter Five: General Discussion

5.1 Summary of results

Synapses are the functional units of the nervous system, mediating the selective processing of information within circuits that leads to the generation of complex behaviours. The incredible complexity of the brain’s structural connectivity, as well as the specificity and plasticity of individual synapses underlies the fidelity and flexibility of neuronal network function. Synaptic specificity is perhaps best known to emerge as a result of the selective interactions of cell adhesion molecules (CAMs) during target selection and synaptic differentiation, however, these interactions alone are not sufficient to promote the formation of connections that are fully functional or capable of plastic remodeling. The selective maturation, retention, and plasticity of individual synapses also requires neurotransmitter-receptor and neurotrophic factor-receptor tyrosine kinase (NTF-

RTK) signaling, transcriptional regulation, and site-specific protein trafficking. These events influence the expression, composition, and localization of various components of pre- and postsynaptic machinery, including neurotransmitter receptors, synaptic vesicles (SVs) and synaptic vesicle proteins (SVPs). Collectively, this differentiates the function of individual synapses, however, the underlying cellular and molecular mechanisms have not been fully defined.

Thus, the main objective of this dissertation was to elucidate the activity- and NTF-dependent mechanisms responsible for establishing pre- and postsynaptic specificity during synaptogenesis.

232

Specific Aim 1: To determine the molecular mechanisms underlying the synaptogenic function

of menin in postsynaptic development.

Conclusions: L-Menin is cleaved at an evolutionarily conserved calpain site in response to NTF-

induced activity. The resulting L-menin proteolytic fragments coordinate nuclear and synaptic

events necessary for postsynaptic development during excitatory cholinergic synaptogenesis

between Lymnaea neurons. This includes (i) subunit-selective nicotinic acetylcholine receptor

(nAChR) gene induction via the N-terminal fragment, and (ii) the postsynaptic clustering of

excitatory nAChRs via the C-terminal fragment, which requires NTF-dependent phosphorylation.

These observations are the first to describe the molecular mechanisms underlying menin’s function

in neurons, and identify a novel synaptogenic mechanism in which a single gene product

coordinates nuclear transcription and postsynaptic targeting of neurotransmitter receptors via

distinct molecular functions of two proteolytic fragments. Furthermore, these findings identify the

C-menin fragment as a candidate molecular scaffold for neuronal nAChR clustering.

Specific Aim 2: To determine whether the molecular actions of menin in neurons are conserved

across evolution.

Conclusions: Mouse menin is cleaved at an evolutionarily conserved calpain site, and the resulting

proteolytic fragments exhibit differential subcellular localization in the central nervous system

(CNS), with menin localized to the nucleus, N-menin to the cytoplasm, and C-menin to the synapse. Menin expression in hippocampal cultures is specifically restricted to neurons, where it mediates subunit-specific transcriptional regulation of nAChR α5. Cholinergic presynaptic

233

facilitation is dependent upon the functional clustering of α7 subunit-containing nAChRs to glutamatergic presynaptic terminals, which likely requires the C-menin proteolytic fragment.

These observations suggest that C-menin may be a molecular scaffold that mediates the clustering of a specific subtype of neuronal nAChRs. Furthermore, MEN1 transcription and menin proteolytic cleavage were disrupted in a mouse model of Alzheimer’s disease (AD), suggesting that menin perturbations may contribute to the emergence of cholinergic dysfunction in AD.

Specific Aim 3: To determine the molecular mechanisms that differentiate the peptidergic composition and function of individual presynaptic terminals.

Conclusions: An identified Lymnaea neuron selectively forms a purely cholinergic synapse or a mixed cholinergic-FMRFamidergic synapse with two distinct postsynaptic targets, and this form of presynaptic transmitter specificity is a plastic phenomenon governed by (i) NTF signaling and

(ii) the target cell-specific production of arachidonic acid metabolites (AA), due to differential G protein coupling of the FMRFamide receptor. The SVP synaptophysin defines the neuropeptide release competency of a synapse, via the selective inhibitory regulation of large dense-core vesicle

(LDCV) release machinery. Synaptophysin gene expression is regulated simultaneously by NTFs and AA metabolites, suggesting that these signaling cascades are contradictory regulators of synaptophysin transcription. Presynaptic transmitter specificity furthermore requires synapse- specific populations of release-competent and -incompetent LDCVs, suggesting that a posttranslational mechanism also acts independent of synaptophysin transcriptional regulation.

These observations identify a target cell-specific retrograde signaling mechanism for presynaptic specificity, via the selective generation of AA, as well as novel form of synaptic plasticity in which

234

the co-transmitter characteristics of individual presynaptic terminals are regulated via NTF-RTK signaling.

235

5.2 Significance of findings, identification of limitations, and future directions

5.2.1 The role of MEN1/menin in postsynaptic specificity

Whereas previous studies have characterized the functional significance of MEN1 expression in synapse formation and plasticity in both invertebrate and mammalian model systems, the molecular mechanisms underlying the effects of menin have remained elusive. In this study, I sought to identify how the synaptogenic effects of menin occur, and found that menin mediates subunit-specific transcriptional regulation of nAChRs to modulate the postsynaptic response.

Perhaps the most surprising discovery I made was that menin also coordinates the functional clustering of nAChRs via a C-terminal proteolytic fragment, generated and targeted by activity- and NTF-dependent signaling.

While the clustering of nAChRs via the C-terminal fragment is conserved from invertebrates to vertebrates, the transcriptional actions of menin appear to have diverged across evolution, with the

N-menin proteolytic fragment localizing to the nucleus and mediating transcriptional regulation in

Lymnaea CNS neurons, whereas full length menin apparently performs this function in mouse hippocampal neurons. One possible explanation for this discrepancy is that the enlarged size of L- menin (84.5 kDa), as compared to mammalian menin (67.5 kDa), may prohibit its interactions with transcriptional regulatory complexes, thus this function may have been adapted by the N- menin fragment in Lymnaea. One of the limitations of this study is that the mechanisms underlying menin’s function in neuronal transcriptional regulation were not pursued. It would be of interest to further dissect how subunit-specific nAChR expression is achieved via menin. For instance, are

236

the transcriptional actions of menin stimulus-specific, dependent upon particular activity

patterns188 or NTFs-RTKs102? Does menin act as a coincidence detector to influence nAChR

subunit expression in response to phosphorylating activation signals induced by activity and RTK

signal transduction? Does transcriptional specificity occur via menin alone, or in a complex with other transcriptional activators, such as MLL1 or SMADs, or transcriptional repressors, such as

JunD or NFκβ111? Are the associations between menin and other transcription factors influenced

by activity- and NTF-dependent signals? Does menin or menin-associated complexes bind

regulatory elements present in the promoters of certain nAChR genes? Future experiments might

include DNA-protein crosslinking, co-immunoprecipitation (co-IP), and transcriptional profiling experiments in combination with specific pharmacological inhibitors, such as those that inhibit transcription factor interactions (e.g. MLL1-menin287), or signal transduction cascades (e.g.

CaMKII or MAPK). These approaches would allow one to determine the molecular associations

amongst menin, other transcription factors, and DNA regulatory elements, whether this is

influenced by activity- and NTF-dependent signals, and the resulting impact on subunit-specific

nAChR gene expression.

Here, I found that the L-menin N-terminal fragment selectively induced transcription of the

cationic L-nAChR C (α-type) subunit, and mouse menin was required for the transcriptional

activation of the nAChR α5 subunit. Phylogenetic analysis reveals that L-nAChR C is closely

related to nAChR α5159, suggesting that context-dependent transcriptional regulation (via

activity/NTF-RTK signaling and menin) of this particular subunit is an evolutionarily conserved

phenomenon. This potentially represents a unique mechanism for modulating neuronal nAChR

function, in which the regulated expression of one gene may be sufficient to alter the functional

237

characteristics of nAChR channels, which is presumably more efficient than having to coordinate

the expression of multiple genes encoding various nAChR subunits. The increased expression of

the nAChR α5 subunit following peripheral nerve injury (which also induces MEN1 expression105),

has been suggested to contribute to the generation of neuropathic pain by potentiating glutamate

release from primary afferent fibers8. Furthermore, mice in which nAChR α5 has been knocked

out have been reported to be resistant to nicotine-induced seizures239, an effect which would also

be consistent with the loss α5-nAChRs potentiating glutamate release at presynaptic terminals.

Together with my findings, these observations support the role of MEN1-dependent subunit-

specific transcription in the regulation of neuronal nAChR function.

One of the limitations of studying neuronal nAChR channels is the difficulty in determining subunit composition and localization, because the antibodies available are highly cross-reactive and/or non-specific. In Lymnaea neurons, I was unable to achieve specific labeling with α-

Bungarotoxin (α-BTX) or various antibodies in vitro or in Western blots (WBs) (data not shown), and opted to design an eGFP-tagged L-nAChR C construct to characterize its synaptic localization.

This method, however, relies on ectopic expression and does not permit investigation of the other nAChR subunits that are present within the channel. Similarly, in mammalian preparations, the precise composition of native neuronal nAChR channels, the subcellular localization patterns of various channel types, and their specific contributions to synaptic function, plasticity, and behaviour have not been unequivocally demonstrated, as much of the available information has been derived from studies in heterologous expression systems. My data from mammalian hippocampal neurons indicate that the C-menin fragment clusters α7 subunit-containing nAChRs to glutamatergic presynaptic terminals, but also that menin selectively mediates expression of the

238

nAChR α5 gene. Considering that menin coordinates nuclear transcription and synaptic clustering

in Lymnaea neurons, it would next be of interest to determine whether presynaptic α7 channels,

which are generally thought to be homopentamers5, might contain α5 as well, at least in regions of

the brain where there is considerable MEN1 expression, such as the hippocampus181, or in response

to NTF-induced MEN1 expression and cholinergic remodeling8,105. Previous indications that α5

modulates glutamatergic presynaptic facilitation (see above), which is normally described as being

dependent upon α-BTX sensitive α7-nAChRs207,226, seem to support this view. α5 has been found to assemble with various other subunits to form functional nAChR channels288, and the existence of α7±α5±β2 heteromeric channels at glutamatergic presynaptic terminals has been proposed289,290,

but molecular validation remains absent. Considering the difficulty of selectively labeling nAChRs

with antibodies in cell culture or WBs, molecular evidence for α7α5 heteromeric channels, and

their localization to glutamatergic presynaptic terminals, might be achieved by fluorescence

resonance energy transfer microscopy via the co-expression of fluorochrome-tagged α7 and α5

subunits in cultured neurons.

The presence of a faster-migrating C-terminal menin immunoreactive band in WBs has been

previously described in a number of mammalian models177,230,231, however, the functional significance of this potential menin fragment had never before been addressed. In this study, I provide the first evidence that the lower C-terminal menin immunoreactive band is a calpain- dependent proteolytic fragment of menin, and that its presence is functionally significant. In

Lymnaea neurons, I found that L-menin is proteolytically cleaved by calpain, and the ~38 kDa C- terminal fragment mediates the postsynaptic clustering of neuronal L-nAChRs in response to activity- and NTF-dependent phosphorylation, likely via CaMKII and MAPK. Furthermore, I

239

found that neuronal α7-nAChRs, both non-synaptic and those clustered at presynaptic terminals,

co-localized with the ~19 kDa C-menin fragment in mouse hippocampal neurons, and that MEN1

knockdown precluded nAChR-dependent presynaptic facilitation. One of the limitations I faced

working with mouse hippocampal cultures was that they did not survive beyond ~3-7 DIV when

exposed to calpain, CaMKII, or MAPK inhibitors (data not shown). This prohibited a more in- depth and comparative investigation of the molecular mechanisms underlying menin’s function in mammalian neurons. For instance, it would have been interesting to determine whether the absence of C-menin, but not menin, as a result of calpain inhibition precludes nicotine-induced presynaptic facilitation without transcriptional perturbations to nAChR α5. The electrophysiological effects of

MEN1 knockdown do, however, suggest that the functional clustering of mammalian neuronal nAChRs occurs via C-menin as well. The observation that the ~19 kDa C-menin immunoreactive fragment was prevalent in excitable tissue such as skeletal muscle and heart provides additional support for the roles of activity, Ca2+ influx, and calpain in the proteolytic cleavage of menin. This finding furthermore suggests that C-menin may influence nAChR clustering in non-neuronal tissues, such that it may be a ubiquitous molecular scaffold or adaptor molecule for nAChR channels, perhaps even playing a previously unrecognized role at the NMJ.

In mouse hippocampal neurons, I found that C-menin co-localized extensively with α-BTX labeled nAChRs, but did not associate strongly with the glutamatergic postsynaptic scaffold PSD-95.

Furthermore, whereas nicotine-induced presynaptic facilitation was eliminated upon MEN1 knockdown, glutamatergic synaptic function was unchanged. Preliminary investigations on neuron cultures derived from a MEN1 conditional knockout (CKO) transgenic mouse by our collaborators in the Netherlands demonstrate that the loss of menin has no effect on the number, structure, or

240

function of autoptic glutamatergic synapses (de Jong PhD thesis; Vrije Universiteit Amsterdam;

unpublished data), and provide support for this view. Considering the well characterized function

of menin as a molecular scaffold111, my observations support the hypothesis that the C-menin fragment is a dedicated scaffold molecule for nAChR clustering. However, further experimental evidence would be required to identify C-menin as the molecular scaffold for neuronal nAChRs.

Thus, the question of whether C-menin is in fact the previously unidentified neuronal nAChR scaffold, or an adaptor molecule that is simply one component of a larger complex remains to be answered. This may be difficult to definitively demonstrate considering that there is likely to be a considerable amount of functional redundancy, given the previously described roles for rapsyn,

PSD-95 and PSD-93 in influencing the clustering of vertebrate neuronal nAChR33,244,245,291. As C- menin/α-BTX puncta were selectively observed at presynaptic terminals, it may be that C-menin is a dedicated scaffold molecule for a particular subtype of neuronal nAChRs. Future experiments using co-IP, mass spectrometry analysis, and selective mutagenesis may pursue direct evidence for the physical association between C-menin and particular nAChR subunits (e.g. α7, α5), and the molecular interactions amongst C-menin and other components of the cholinergic postsynaptic density (PSD) at glutamatergic presynaptic terminals.

It would next be of interest to determine the mechanism by which C-menin mediates the functional clustering of nAChR. For instance, how do the molecular interactions between C-menin and nAChRs occur? Given the size difference between Lymnaea and mouse C-menin, are the structural associations different, or is there a conserved common functional element? Is the clustering and synaptic targeting of nAChRs dependent upon C-menin phosphorylation, and if so, at which residues and by what kinases? Are these actions tied to a specific set of NTF-RTKs? Is the

241

functional clustering of neuronal nAChRs by C-menin subtype specific, and if so, how is this selectivity achieved? As human menin is phosphorylated at two conserved C-terminal serine residues204, it would be of interest to determine whether selective mutagenesis of these particular

residues might be sufficient to abolish C-menin self-association, synaptic localization, or nAChR clustering. Considering that there are (at least) two residues of C-menin that are phosphorylated, it is an intriguing hypothesis that C-menin/nAChR clustering may be potentiated by a coincidence detection mechanism that requires phosphorylation at distinct sites by different kinases activated in response to activity and NTF-RTK signaling. Selective mutagenesis and serial deletion studies of GFP-tagged mouse rapsyn co-expressed with muscle nAChR subunits in HEK-293 cells have determined that nAChR clustering occurs through a C-terminal coiled-coil domain, rapsyn self- association occurs through medial tetratrichopeptide repeats, and membrane targeting occurs through the 15 N-terminal residues32. Future studies might employ a similar strategy with the C- menin fragment to determine whether the molecular mechanisms underlying nAChR clustering via C-menin are similar. Preliminary structural modeling indicates that Lymnaea C-menin likely contains a coiled-coil domain, whereas mouse C-menin may not (Fig. 5.1; SIB-ExPASy,

COILS)32. Perhaps the expanded Lymnaea C-menin sequence is sufficient to mediate neuronal

nAChR clustering autonomously in a manner similar to rapsyn, whereas another molecular co- factor may be required for neuronal nAChR clustering by C-menin in mammalian neurons. An intriguing candidate co-factor is β-catenin, another dual-function protein that mediates cell-cell adhesion events and transcriptional regulation in response to Wnt signaling. β-catenin is furthermore known to (i) interact with menin180, (ii) play a variety of roles at the synapse, including nAChR clustering at the NMJ292, and (iii) contains a predicted coiled-coil domain (Fig. 5.1).

Alternatively, it may be that the clustering of neuronal nAChRs occurs through a mechanism that

242

is distinct from rapsyn-mediated clustering, but shared by menin orthologues – perhaps the

presence of a novel functional domain in the ~50 C-terminal residues would offer an explanation

for its remarkable conservation across evolution (see Fig. 2.3).

One of the limitations of this study is that I was unable to differentiate the contributions of menin,

via the transcriptional regulation of specific nAChR subunits and the synaptic clustering of nAChR

channels, to cholinergic synaptic function. Previous studies from our lab have demonstrated the

functional necessity99 and sufficiency102 for L-MEN1 in excitatory cholinergic synaptogenesis

between Lymnaea neurons. However, the question of whether either component of this coordinated

mechanism is sufficient and/or required for cholinergic postsynaptic function remains to be

definitively answered. One possibility is that transcriptional regulation of L-nAChR C by L-menin

or nAChR α5 by menin is sufficient for cholinergic function because the clustering function of C-

menin is redundant, perhaps due to the compensatory actions of rapsyn in neurons. In chick ciliary

ganglion neurons, the internal loop of the α3 subunit has been shown to mediate the postsynaptic

targeting of neuronal nAChRs293, suggesting that specific subunits influence the targeting of distinct receptor subtypes to particular subdomains. Thus, transcriptional regulation of L-nAChR

C or nAChR α5 subunits by menin may be sufficient to direct cholinergic postsynaptic function via inducing the expression of nAChR channels that target to appropriate synaptic sites. Another, perhaps more likely, possibility is that C-menin clustering of neuronal nAChRs is sufficient in the absence of transcriptional regulation of L-nAChR C or nAChR α5. For instance, α5 knockout mice exhibit otherwise normal nAChR expression and clustering, brain development and cognitive behaviour239, such that the expression of particular subunits may not be an absolute requirement.

My observation that MEN1 knockdown in mouse hippocampal neurons fully eliminated nicotine-

243

induced presynaptic facilitation provides strong support for the notion that C-menin-dependent clustering is required for neuronal nAChR function, as no other perturbations of specific nAChR subunits (e.g. α5294, α7236) or proposed molecular scaffolds (e.g. rapsyn33, PSD-93245) achieve a comparable loss of function upon knockout. Finally, a third possibility that should be considered is that both of these processes may ultimately be necessary for MEN1-dependent cholinergic synaptogenesis, and the proper development, function, and maintenance of cholinergic synapses in the CNS.

Future experiments designed to dissect the contributions of menin to transcriptional regulation and nAChR clustering may be difficult to achieve in the Lymnaea model system, because of (i) the limited ability to perform genetic manipulations, and (ii) the apparent requirement for activity- and

NTF-dependent L-menin proteolytic cleavage and C-menin phosphorylation, signaling events that would also induce endogenous L-MEN1. There are, however, a number of experimental options to circumvent this dilemma. For instance, RNA interference (RNAi)-mediated knockdown of L- nAChR C would permit one to ask whether subunit-specific transcriptional regulation is required for L-MEN1-dependent excitatory cholinergic synaptogenesis, whereas RNAi-mediated knockdown of L-MEN1 with mRNA injection of L-nAChR C would permit one to ask whether

MEN1 expression is required. Designing C-menin phosphomimetic mutants may also permit one to ask whether C-menin is sufficient to promote excitatory cholinergic synaptogenesis in the absence of NTF-RTK signaling, endogenous L-MEN1 expression and L-nAChR C transcriptional activation. These questions may be simpler to address with the MEN1 CKO mouse. For example, using hippocampal cell cultures from this line, the endogenous MEN1 gene could be knocked-out, and gain of function experiments could be performed by viral knock-in of (i) the C-menin

244

fragment, or (ii) full length MEN1 in which the calpain cleavage site has been mutated. This would

facilitate experiments in which the necessity and sufficiency of menin in transcriptional regulation

and C-menin in nAChR clustering could be isolated in the context of a null background.

In contrast to the specific clustering actions of the C-menin fragment, the transcriptional actions

of menin in the regulation of synaptic function may be comparatively non-specific. For instance, despite the subunit-specific reduction in nAChR α5, I found that MEN1 knockdown in mouse hippocampal neurons induced an apparently global transcriptional upregulation. This observation suggests that changes in MEN1 expression may have wide-spread effects on synaptic function that

are independent of its regulation of nAChR channel subunit composition and localization. For

example, peripheral nerve injury-induced MEN1 upregulation in mice has been found to decrease

the expression of glutamic acid decarboxylase 65 (GAD65), the enzyme which catalyzes glutamate

conversion to GABA, thereby contributing to neuropathic pain by shifting the balance of synaptic

transmission in favor of glutamate107. It is, however, intriguing that this represents a distinct

mechanism to achieve potentiation of glutamatergic presynaptic function which is also coordinated

by menin. A previous study from our lab demonstrated that RNAi-mediated L-MEN1 knockdown simultaneously in pre- and post-synaptic Lymnaea neurons inhibited inhibitory and excitatory

dopaminergic synapse formation99. Although the synaptogenic mechanism at these non-

cholinergic synapses was not investigated further, I anticipate that it involves transcriptional perturbations of other synapse-associated genes, perhaps even on the presynaptic side, rather than the postsynaptic actions of the C-menin fragment. We have also previously found that NTF-

induced bursting activity is not a universal mechanism in Lymnaea neurons, but rather occurs in a

cell type-specific manner101. This suggests that regulated L-MEN1 expression and L-menin

245

proteolytic cleavage would also occur in a cell type-specific manner, and thus potentially identifies a novel mechanism for synaptic specificity in which NTF-dependent excitatory cholinergic synaptogenesis may be selectively governed via MEN1 gene induction of and the generation of C- menin. It would be of interest to determine whether L-MEN1 expression is induced by NTF-RTK signaling in neurons other than LPeD1, for instance, those that do not exhibit NTF-induced bursting, and whether L-menin is proteolytically cleaved in these neurons. This potential cell type- dependent mechanism for synaptic specificity has interesting implications for the regulation of cholinergic network function, and will be discussed in further detail below (Section 5.2.3).

246

Figure 5.1 – A potential mechanism for neuronal nAChR clustering by coiled-coil domains

Rapsyn-dependent clustering of nAChRs at the NMJ occurs via a coiled-coil domain (top left;

SIB-ExPASy COILS prediction). The Lymnaea C-menin fragment similarly contains a predicted

coiled-coil domain (top right), suggesting that Lymnaea C-menin may cluster neuronal nAChRs

by a similar protein domain. The prediction score for mouse C-menin is much lower (bottom right), thus there may be a distinct mechanism for neuronal nAChR clustering, or it may occur through a co-factor such as β-catenin, which contains a predicted coiled-coil domain (bottom left).

247

5.2.2 The role of neuropeptides and synaptophysin in presynaptic specificity

The synaptic regulation of peptide neurotransmitters is perhaps the one area of neuroscience we know the least about. Invertebrate models are valuable tools for the study of peptidergic synaptic transmission because they are one of the few systems in which measurable synaptic responses can be directly attributed to the actions of identified neuropeptides. In this study, I took advantage of observations that invertebrate co-transmitting neurons vary the use of classical and peptide neurotransmitters according to the identity of the postsynaptic target, and sought to identify the molecular mechanisms underlying (i) the regulation of peptidergic synaptic transmission and (ii) the emergence of presynaptic transmitter specificity. Here, I identified AA metabolites as a target cell-specific retrograde signaling mechanism for presynaptic specificity, a novel role for synaptophysin in the inhibitory regulation of LDCV release, as well as a new component of NTF- dependent synaptic plasticity in which the co-transmitter characteristics of individual presynaptic terminals are regulated.

Previous studies of co-transmission in invertebrate models have found that synaptic transmission between a single presynaptic neuron and distinct postsynaptic targets can be differentially mediated by classical and peptide neurotransmitters, independent of variations in postsynaptic receptor expression137-139. However, whether this type of presynaptic specificity involves the selective localization or release of SSVs and LDCVs at distinct presynaptic terminals remained undetermined. Here, I found that a co-transmitting neuron released FMRFamide, but not ACh, in a synapse-specific manner. FMRFamide neuropeptide localization was found to be enhanced at presynaptic sites in synaptic pairs exhibiting purely cholinergic synaptic transmission, which

248

implicates a mechanism for the inhibition of LDCV release at particular presynaptic terminals.

Surprisingly, the selective regulation of presynaptic properties was not solely determined by a target cell-dependent retrograde signal, as had been anticipated by previous studies128,129,131,132.

Rather, I found that presynaptic specificity was a plastic phenomenon governed by NTF-RTK signaling, as target cell-dependent transmitter specificity was reversed in the presence or absence of NTF. This finding identifies a new mechanism of action for NTFs in the regulation of presynaptic function during synapse formation, maturation, and plasticity. One of the limitations of this study is that the identity of the NTF and RTK mediating this response remains unknown.

While the use of CNS-conditioned media offers the advantage of recapitulating the in vivo environment of extrinsic trophic factors, it is compositionally uncharacterized, and is known to contain a variety of NTFs with a wide range of synaptic influence. Fractionation analysis of

Lymnaea CM by our group identified a cysteine-rich NTF (CRNF) with effects on neurite outgrowth and Ca2+ currents295. Previous studies from our group have also characterized the presence of L-EGF in CM, and its role in promoting excitatory cholinergic synapse formation via

L-EGFR100,296. Furthermore, neurotrophins and Trk receptors have recently been identified in

Aplysia297, and a neurotrophin signaling system is thus likely to be present in Lymnaea as well.

Considering that (i) specific NTFs and RTKs are known to differentially instruct presynaptic maturation (e.g. glutamatergic and GABAergic specificity via FGF isoforms116), and (ii) spatially segregated axon terminals of a single presynaptic neuron are likely to encounter different profiles of NTFs due to regional expression patterns, it is of interest to pursue identification of the relevant

NTFs and RTKs, as well as the downstream signaling mechanisms mediating this functional differentiation. Future studies might employ cell culture approaches with purified NTFs or RNAi- mediated knockdown of specific RTKs, in combination with specific pharmacological inhibitors

249

(e.g. MAPK), approaches we have previously used to dissect L-EGF- and L-EGFR-dependent signaling mechanisms in excitatory cholinergic synaptogenesis100,102.

Considering that FMRFamide accumulated at non-releasing synaptic sites, and that LDCVs are

synthesized from the endoplasmic reticulum/Golgi network, trafficked to synapses, and do not

recycle, the transcriptional regulation and selective actions SVPs was an intriguing candidate

mechanism to account for the synapse-specific facilitation and inhibition of FMRFamide release.

This is supported by the observation that co-transmitting neurons formed cholinergic synapses regardless of postsynaptic target identity, such that the cytoplasmic matrix of the active zone

(CAZ) must be comparatively unchanged. Here, I found that FMRFamidergic synaptic transmission induces synthesis of the retrograde transmitter AA when it generates an excitatory

2+ postsynaptic response via Gαi/o-linked G protein coupled receptors (GPCRs), leading to the Ca -

and Gβγ-dependent activation of PLA2. AA acts trans-synaptically to tune the neuropeptide release competency of the innervating terminal by transcriptional regulation of synaptophysin.

NTF signaling upregulated synaptophysin expression, and this was counteracted by AA signaling.

Furthermore, in the absence of NTFs, I found that AA upregulated synaptophysin expression, suggesting that AA-evoked cascades mediate both transcriptional activation and repression according to the context of NTF-evoked signaling. This observation is the first to extend the effects

of AA metabolites on synaptic plasticity171 to include trans-synaptic transcriptional regulation. It

would next be of interest to dissect the mechanisms by which AA and NTF regulate synaptophysin

gene induction, and how these signals are counteracted. AA has previously been found to activate

PKC172 in rat synaptosomal membranes, and inhibit presynaptic Kv channels at mouse mossy fiber synapses resulting in broadening of the action potential and thus enhanced Ca2+ influx4. Both of

250

these molecular signaling events are well known to influence transcription, and often involve

activation of the transcription factor NFκβ298. Future experiments might include ectopic AA at synapses with inhibitory FMRFamidergic transmission (e.g. VD4-RPeD1) to further validate the requirement for target cell-specific induction of AA in presynaptic transmitter specificity.

Pharmacological activation or inhibition of PKC, or RNAi-mediated knockdown of NFκβ, could also be performed in combination with synaptophysin gene expression profiling and electrophysiological analysis of synaptic transmitter characteristics to pursue identification of the molecular mechanisms underlying bi-directional transcriptional regulation.

Here, I found that high levels of synaptophysin expression inhibited neuropeptide release, whereas low levels were facilitative. Intriguingly, this effect was specific to LDCV release, as cholinergic synaptic transmission via small synaptic vesicle (SSV) release was unaffected by manipulations of synaptophysin expression. Furthermore, this outcome was not paralleled by manipulations of other SVPs such as two functionally distinct synaptotagmin isoforms. Another notable finding from this study was that synaptotagmin I C2B-α selectively inhibited SSV, but not LDCV release.

One question that remains to be answered is how the functional specificity of SVPs is achieved amongst SSVs and LDCVs to selectively regulate the release of co-transmitter substances. It is likely that the answer lies in the differences of SVP composition between SSVs and LDCVs, as well as the distinct mechanisms of SV biogenesis, docking, and release. As synaptophysin is generally found to be more concentrated in SSVs21, one possibility is that the phosphorylation-

dependent processes of SSV docking and/or membrane recycling also induce a posttranslational

modification (PTM) that facilitates dissociation of synaptobrevin from synaptophysin, and thus

selectively eliminates inhibitory regulation of SNARE complex assembly in recycling SSVs18,279.

251

Future experiments might include mass spectrometry analysis and mobility shift assays to

determine whether there are different populations of glycosylated and/or phosphorylated

synaptophysin within neurons. If target residues for PTMs are identified, mutagenesis could be

performed to determine whether particular PTMs selectively influence SSVs and LDCVs release

characteristics. The generation of phospho-specific or glycosylation-specific antibodies could also

be used in combination with electron microscopy to determine whether SSVs and LDCVs exhibit

distinct profiles of synaptophysin PTMs.

While I found transcriptional regulation of synaptophysin to influence neuropeptide release,

presynaptic transmitter specificity was also maintained irrespective of synaptophysin expression

levels in the presence of competing molecular signals from distinct postsynaptic targets. This

observation identifies the requirement for an as-yet unidentified mechanism that generates synapse-specific populations of LDCVs with distinct release characteristics. Here, I have proposed two putative scenarios for synapse-specific facilitation and inhibition of neuropeptide release: (i) synaptophysin is selectively incorporated or excluded from LDCVs, and different populations of vesicles are in turn selectively trafficked to individual presynaptic terminals; or (ii) uniform populations of LDCVs are trafficked to presynaptic terminals, and synapse-delimited PTM of

synaptophysin modulates LDCV release capacity. One of the limitations of this study is that I was

unable to validate expression changes of synaptophysin and the composition of LDCVs. Protein-

level validation is typically one of the shortcomings of the Lymnaea model system, as evolutionary

sequence divergence often precludes the use of commercial antibodies. In future experiments, this

could be circumvented by the use of an epitope-tagged construct, although a caveat of this

approach is that it relies on ectopic expression. Another option would be to generate a Lymnaea

252

synaptophysin-specific antibody. Electron microscopy would then allow one to determine whether

LDCV synaptophysin immunoreactivity is distinct or uniform at presynaptic terminals with

different postsynaptic targets, and thus obtain experimental support for one of the above

mechanisms. Once a candidate pathway is identified, it would be of interest to delineate how this

is differentially influenced by AA and NTF signaling. Previous studies on the regulation of

synaptophysin vesicular sorting and SVP interactions suggest that a potential candidate mechanism

for the differential sorting of synaptophysin to LDCVs is N-glycosylation272, whereas synapse- delimited PTM may be regulated via kinase crosstalk in the presynaptic terminal, perhaps via

GSK-3β273 and PKC172, to influence synaptobrevin association and inhibitory regulation of

SNARE complex formation. Finally, the possibility that a third scenario in which synapse-specific

facilitation and inhibition of neuropeptide release arises from an alternative, synaptophysin- independent mechanism should also be considered. However, as synaptophysin fully accounted for neuropeptide release competency in monosynaptic culture, it seems unlikely that neurons would recruit an entirely different mechanism for this purpose in the context of polysynaptic innervation.

Neuropeptides are expressed in neurons prior to synaptogenesis67,68, however, their contributions

to this process have remained largely undefined. In this study, I have shown that differential

coupling of the FMRFamide metabotropic receptor to distinct molecular signaling cascades

influences presynaptic maturation to differentiate the functional characteristics of individual

presynaptic terminals. Previous observations that FMRFamide signaling inhibits neurotransmitter release machinery168, and that this action controls the timing and directionality of synaptogenesis170, suggest that neuropeptide signaling is an important instructive signal for the

253

assembly, maturation, refinement, and maintenance of synaptic networks. This concept is

furthermore supported by the observation that some neurons exhibit transient FMRFamide

expression during nervous system development69. An interesting avenue to pursue would be to

determine whether peptidergic signaling is required for appropriate CNS development and

function. Future experiments may include global knockdown, neuron-specific knockdown, or

ectopic expression of FMRFamide gene transcripts, combined with analyses of the effects on

structural connectivity, gene expression, electrophysiological function, and behaviour. My present

observations that the emergence of presynaptic transmitter specificity during synaptic maturation

occurs via the target cell-specific production of AA metabolites furthermore identifies a previously

unrecognized role for this signaling system in nervous system development. Importantly, it seems

that this retrograde transmitter is also likely to underlie the differentiation of other forms of target

cell-dependent presynaptic specificity as well. For instance, the specific induction of long term

potentiation or depression at individual presynaptic terminals innervating mammalian

hippocampal pyramidal and interneurons, respectively, has been reported to involve postsynaptic

Ca2+ elevation and presynaptic PKC signaling134-136, both of which are tied to the generation and

signal transduction of AA metabolites. It would next be of interest to determine whether the target

cell-specific generation of AA is a universal mechanism for presynaptic specificity, one that is dependent not upon the identity of a particular neurotransmitter, but the differential G protein coupling of metabotropic receptors in distinct postsynaptic targets. The potential implications for this mechanism in the regulation of neuronal network function will be discussed in further detail below (Section 5.2.3).

254

5.2.3 Implications for synaptic network function

5.2.3.1 Modeling behaviour in Lymnaea by the synaptic properties of individual neurons

The Lymnaea model system offers a unique opportunity to manipulate a single neuron in vivo and observe the behavioural effects of disrupting a simple synaptic network145. While determining the

mechanisms underlying the transcriptional regulation of MEN1 was not a primary objective of this

study, the possibility that NTF-induced expression may occur in a cell type-specific manner has

interesting implications for menin in synaptic specificity and the regulation of neuronal network

function. The postsynaptic properties of LPeD1, a cardioexcitatory neuron, exhibit a NTF- and L-

MEN1-dependent switch from inhibitory to excitatory cholinergic transmission. RPeD1, by

contrast, maintains an inhibitory postsynaptic profile in the presence or absence of NTFs. This

suggests that L-MEN1 is not similarly induced by NTFs in RPeD1, and is supported by

observations that human MEN1 transcripts exhibit 5’ untranslated region variation due to six

distinct transcription start sites that are influenced by multiple regulatory elements with cell type-

specific and context-dependent activity199. The maintenance of an inhibitory cholinergic connection between VD4-RPeD1 is essential for respiratory behaviour, as reciprocal inhibition forms the basis of the respiratory central pattern generator (rCPG) which drives antagonistic motor functions, leading to pneumostome opening (expiration) and closing (inspiration)144 This behaviour can be selectively ablated, for instance, by the removal of VD4145. While the

postsynaptic cholinergic properties of RPeD1 are not influenced by NTF signaling, I found that

the presynaptic peptidergic characteristics of VD4 exhibit a NTF-dependent switch due to the transcriptional regulation of synaptophysin. In the absence of NTFs or synaptophysin,

FMRFamidergic transmission between VD4-RPeD1 may inhibit dopamine release from RPeD1

255

and disrupt rCPG rhythmogenesis, thus precluding respiratory behaviour. NTF-dependent synaptic

network plasticity is the model proposed by our group for the bidirectional regulation of heart rate

and breathing behaviour during periods of hibernation and activity in Lymnaea. Future studies in

the intact animal could next be pursued to determine the functional significance of the novel

activity- and NTF-dependent mechanisms for pre- and postsynaptic specificity that I have

identified. For instance, would RNAi-mediated knockdown of synaptophysin in VD4 facilitate

FMRFamide release, thereby precluding the VD4-RPeD1 reciprocal synapse, rCPG activity, and breathing behaviour? Would ectopic AA disrupt presynaptic transmitter specificity and similarly preclude breathing? Would ectopic L-MEN1 expression in RPeD1 via mRNA microinjection lead to the formation of an excitatory cholinergic synapse between VD4-RPeD1, and disrupt rCPG network activity and breathing?

5.2.3.2 Potential applications to human health and disease

Cholinergic transmission plays a critical role in brain development, function, and plasticity. In the hippocampus, activation of nAChRs is known to regulate the expression of NTFs250, promote

synaptic plasticity, and also the survival, maturation, and synaptic integration of adult-born neurons247. The hippocampus is the main target of cholinergic innervations, and the highest levels

of MEN1 expression are found there181. In addition to the β-amyloid/tau plaque hypothesis, AD pathophysiology is characterized by hippocampal pathology associated with reduced NTF expression210, an extensive loss of cholinergic function218, and synaptic failure299. β-amyloid oligomers are furthermore known to both inhibit and inappropriately activate nAChRs300,301. An

emerging hypothesis in the field is that tumor suppressor dysfunction underlies the emergence of

256

synaptic dysfunction in AD, however, the tumor suppressors identified to date (BRCA1110,302 and

PTEN109) have not been specifically linked to cholinergic deficits. My observations indicate that

MEN1 disruption precludes nAChR-induced synaptic plasticity, and that MEN1 gene function is perturbed in a mouse model of AD. Could perturbations of NTF-RTK signaling, MEN1 expression,

C-menin cleavage, and the functional expression and clustering of nAChRs be a causative

mechanism for cholinergic loss in AD? Could promoting MEN1 gene function maintain

cholinergic synapses and alter AD progression? A proposed model for the role of MEN1 in AD is

presented in Fig. 5.2.

The mechanism of action underlying the functional significance of the synaptophysin family of

SVPs is currently unknown. Simultaneous knockout of synaptophysin and synaptogyrin has been

found to induce specific deficits in synaptic plasticity19. Synaptophysin is also one of the genes

associated with X-linked intellectual disability285. My observations indicate that synaptophysin is

a critical regulator of peptidergic synaptic transmission, influenced simultaneously by AA and

NTF signaling. Considering that neuropeptides are important modulatory signals for the synaptic

plasticity events underlying learning and memory25, it seems likely that the regulation of peptide

neurotransmitters is responsible for synaptophysin-dependent synaptic plasticity. Neuropeptide Y

(NPY) is the most prevalent neuropeptide in the brain, and perturbation of NPY-dependent synaptic plasticity has been implicated in the etiology of neuropsychiatric conditions such as depression and schizophrenia303. These conditions are also known to be influenced by NTF

deficits283,284. Could synaptic plasticity dysfunction in psychiatric disorders result from perturbed

NTF signaling, synaptophysin expression, and the inappropriate induction of peptidergic synaptic

257

transmission? A proposed model for the role of NTFs and synaptophysin in psychiatric disorders is presented in Fig. 5.2.

258

Figure 5.2 - Proposed model for the clinical relevance of synaptic specificity mechanisms

259

5.2.4 General conclusion

Defining the fundamental mechanisms that determine neuronal network assembly and function is

critical for our understanding of the brain, however, its incredible complexity often still

necessitates the use of reductionist approaches or simple model systems for experimental

investigations. In this dissertation, I sought to define the cellular and molecular mechanisms

underlying synaptic specificity, focusing on two significant questions: (i) how does a neuron

selectively generate appropriate and diverse types of presynaptic outputs or postsynaptic

responses, and (ii) how does a neuron establish functionally distinct synapses with specific

postsynaptic targets? Here, I have identified novel roles for activity- and NTF-dependent signaling in differentiating pre- and postsynaptic function during synapse formation and maturation, and furthermore provided the first ever insights into two often observed but mechanistically

uncharacterized phenomena: (i) the synaptogenic actions of the MEN1 tumor suppressor gene; and

(ii) presynaptic transmitter specificity and the regulation of peptidergic synaptic transmission. The

findings presented here identify new avenues for future investigations by defining tractable cellular

and molecular mechanisms underlying activity- and NTF-dependent synaptic specificity. Moving forward, there is still much to be learned about how synaptic specificity in the CNS emerges via the selective formation, refinement, and differentiation of individual synapses to achieve the magnificent functional capabilities of the brain.

260

Chapter Six: Appendix I: Overview of Lymnaea experimental preparations

6.1 Synaptic cell culture

The procedure for the isolation and culture of Lymnaea CNS neurons has been described in detail

in the following book chapter:

Syed NI, Zaidi H & Lovell P. (1999). In vitro reconstruction of neuronal circuits: a simple model system approach. Modern techniques in neuroscience research (pp. 361-377). Springer- Verlag Berlin Heidelberg. ISBN: 978-3-642-58552-4.

Table 6.1 - Identified neurons employed in this thesis for the study of synaptic specificity

Neuron Neurotransmitter Phenotype Synaptic Characteristics Visceral Dorsal 4 Acetylcholine (ACh), Central interneuron (VD4) FMRFamide (presynaptic) Left Pedal Dorsal 1 Serotonin (5-HT) VD4→LPeD1 – excitatory (LPeD1) cholinergic (postsynaptic) Right Pedal Dorsal 1 Dopamine VD4↔RPeD1 – inhibitory (RPeD1) cholinergic; inhibitory dopaminergic (bidirectional) Visceral F group neurons FMRFamide VD4→VF – excitatory mixed (VF) (postsynaptic) Pedal A cluster neurons 5-HT PeA↔PeA – electrical synapse (PeA) Cerebral Caudodorsal Egg Laying Hormone (ELH) Neuroendocrine cells Cells (CDC)

261

Figure 6.1 - Lymnaea cell culture preparations

262

Figure 6.1: Lymnaea cell culture preparations

(A). Schematic of the Lymnaea CNS, depicting the locations of the neurons used that were

employed in this dissertation. Dorsal surface of the central ring ganglia is shown. L/RCe, left and

right cerebral ganglia; L/RPe, left and right pedal ganglia; L/RPl, left and right pleural ganglia;

L/RP, left and right parietal ganglia; V, visceral ganglion. (B). Schematic of the approach for

isolation of identified neurons from trypsinized CNS using suction applied through a glass pipette.

Isolated neurons are then deposited from the pipette onto poly-L-lysine coated glass coverslips

(not shown). (C). Depictions of identified Lymnaea neurons in culture. Representative images demonstrate the configuration of soma-soma, triple-soma and axon-axon paired neurons used for the study of synaptic specificity. Single cell (unpaired, not shown) and soma pairing was used for electrophysiological and single cell qPCR approaches; axon pairing was used to resolve the synaptic sites for immunocytochemical approaches. Scale bar, 20 μm.

263

6.2 Single cell qPCR

To characterize the gene expression changes of nAChR subunits induced by NTF signaling and

MEN1 expression in postsynaptic LPeD1 neurons (Chapter 2), or to characterize the gene expression changes of synaptic vesicle proteins induced by target cell-dependent neurotransmitter- receptor interactions (Chapter 4), I opted to develop a single cell qPCR approach for Lymnaea neurons. This technique was adapted from a method recently described in Nature Protocols, has been described in detail in a recent publication from our lab, and is described briefly below:

Citri A, Pang ZP, Südhof T, Wernig M & Malenka RC. (2012). Comprehensive qPCR profiling of gene expression in single neuronal cells. Nat. Protoc. 7(1), 118-27. Flynn N, Getz A, Visser F, Janes TA & Syed NI. (2014). Menin: a tumor suppressor that mediates postsynaptic receptor expression and synaptogenesis between central neurons of Lymnaea stagnalis. PLoS One, 9(10), e111103.

1. Cultured neurons were perfused with sterile Lymnaea saline prior to sample collection.

2. The cytoplasm of the target cultured neuron was isolated using suction applied through an

autoclaved patch electrode, deposited into RT-PCR reaction buffer, and then immediately

frozen on dry ice. Because the genome of Lymnaea has not been sequenced, I was unable to

design intron spanning primers to ensure qPCR amplified products were derived from mRNA

transcripts and not DNA. Therefore, isolating the cytoplasm and leaving the nucleus in culture

circumvented this limitation.

3. RT-PCR amplification (35 cycles) of the genes of interest (GOIs) was performed with the

SuperScript III one-step RT-PCR system with Platinum Taq DNA polymerase (Invitrogen),

using a gene specific primer (GSP) mixture that amplified a large section of the coding region 264

for the intended targets (set 1: housekeeping genes 18s rRNA and β-tubulin, L-MEN1 and L-

nAChR subunits A-L; set 2: 18s rRNA and β-tubulin, 4/7-FMRFamide, synaptic vesicle

proteins Syp, Syt-α/β, SYX, SYB, and G proteins αi, αo, αq, αs, αa and β).

4. PCR cleanup was performed with ExoSAP-IT (Affymetrix) to eliminate GSPs and dNTPs.

5. Single cell samples (n=6-15) were pooled into 3 fractions, and qPCR amplification (45 cycles;

3x triplicate replicates) was performed with the QuantiTect SYBR green PCR kit (Qiagen)

using GSPs directed to a target-specific region of 80-120 base pairs. Efficiency values for each

primer set were verified by serial dilution curves, and ranged between 80-120%, R-squared

coefficient of regression values ranged between 0.94-1.00, and all products were found to

amplify within the linear dynamic range determined for each primer set. Melting curves were

assessed to ensure all CT values included in analyses were the result of target specific amplified

products. Identity of the amplicons was verified by DNA gel electrophoresis to ensure they

were the correct size, and then DNA sequencing was performed on each amplicon using the

gel-purified samples.

6. Negative control reactions included: (i) no template, (ii) external solution (Lymnaea saline)

(samples were treated as above), and (iii) cDNA-negative cell cytoplasm (samples were treated

as above only without reverse transcription). This was performed for each primer set, and

reactions were found to be mostly negative. However, in negative control and experimental

samples PCR products were occasionally observed to amplify beyond the linear dynamic range

(after ~40 cycles) and were discernable as non-specific by distinct melting curve values. These

CT values were excluded from experimental analysis.

265

Figure 6.2 - Lymnaea single cell qPCR

266

Figure 6.2: Lymnaea single cell qPCR

(A). Cell cytoplasm approach for unpaired (top) or soma-soma paired (bottom) neurons. Scale bar,

20 μm. (B). Deposition of cytoplasm samples into RT-PCR reaction buffer. (C). Design of GSPs for RT-PCR and qPCR. (D). qPCR amplification of GOIs in single cells. For example, expression of the reference gene β-tubulin is within ~1-2 cycles amongst LPeD1 neuron samples, whereas L-

MEN1 expression is upregulated in CM-treated LPeD1 neurons, but below the detection threshold in DM-treated LPeD1 neurons. (E). Uniform melting curves verify that the products amplified during the qPCR reactions are specific and identical. (F). DNA gel electrophoresis ensures that qPCR amplified products are the appropriate size (target size 80-120 base pairs; lower markers represent 50 and 150 base pairs). For example, variations in band intensity is indicative of differences in the relative expression levels of housekeeping genes, L-MEN1 and L-nAChR subunits within LPeD1 neurons. (G). DNA sequence analysis verifies that the intended targets are amplified during the qPCR reaction.

267

Chapter Seven: Appendix II: Acknowledgements and contributions

Chapter 2: Two proteolytic fragments of menin coordinate the nuclear transcription and postsynaptic clustering of neurotransmitter receptors during synaptogenesis between

Lymnaea neurons (this manuscript has been published under an open access license)

Collaborators: Angela M. Getz, Frank Visser, Erin M. Bell, Fenglian Xu, Nichole M. Flynn, Wali

Zaidi, Naweed I. Syed.

Contributions: Experimental design: AG, FV, NS. Manuscript and figure preparation: AG, EB

(Fig 2.11). Cell culture and tissue preparation: AG, WZ (synaptic and single cell culture), NF

(single cell culture for qPCR - partial). Electrophysiology: AG, FX (single cell recordings - partial). Molecular biology: AG. ICC: AG. Imaging: AG, EB (epitope tags imaging - partial).

Funding: This work was supported by a Canadian Institutes of Health Research (CIHR) grant to

NS. FX was funded by a Natural Sciences and Engineering Research Council of Canada (NSERC) grant. AG and NF were funded by Alberta Innovates – Health Solutions (AIHS) and NSERC studentships. EB was supported by an NSERC undergraduate student research award. This work was supported by the Molecular Biology Core Facility and the Regeneration Unit in Neurobiology

(RUN) Advanced Optical Microscopy Imaging Core Facility of the Hotchkiss Brain Institute at the University of Calgary (Calgary, Canada).

Competing interests: None.

268

269

270

271

272

273

Dr. Frank Visser (University of Calgary) is a co-author on the following manuscript:

Getz AM, Visser F, Bell EM, Xu F, Flynn NM, Zaidi W & Syed NI (2016). Two proteolytic fragments of menin coordinate the nuclear transcription and postsynaptic clustering of neurotransmitter receptors during synaptogenesis between Lymnaea neurons. Scientific Reports. 6, 31779; doi 10.1038/srep31779.

This manuscript has been reproduced, in part, in the Ph.D. thesis of Angela Michelle Getz, in the

Department of Neuroscience, Faculty of Graduate Studies, University of Calgary.

The thesis will be published in the Library and Archives Canada (through the University of

Calgary). To allow this, Angela Michelle Getz has signed a Theses Non-Exclusive License that authorizes Library and Archives Canada to reproduce, communicate to the public on the Internet, loan, distribute or sell copies of the above listed thesis, among other things. By signing this document, I, Frank Visser, indicate that I have read and understand this agreement and I allow the listed work to be published by Library and Archives Canada.

(Signature removed)

______

Dr. Frank Visser

November 14, 2016

______

Date

274

Erin Bell (University of Calgary) is a co-author on the following manuscript:

This manuscript has been reproduced, in part, in the Ph.D. thesis of Angela Michelle Getz, in the

Department of Neuroscience, Faculty of Graduate Studies, University of Calgary.

The thesis will be published in the Library and Archives Canada (through the University of

Calgary). To allow this, Angela Michelle Getz has signed a Theses Non-Exclusive License that authorizes Library and Archives Canada to reproduce, communicate to the public on the Internet, loan, distribute or sell copies of the above listed thesis, among other things. By signing this document, I, Erin Bell, indicate that I have read and understand this agreement and I allow the listed work to be published by Library and Archives Canada.

(Signature removed)

______

Erin Bell

September 14, 2016

______

Date

275

Dr. Fenglian Xu (University of Calgary, Saint Louis University) is a co-author on the following manuscript:

This manuscript has been reproduced, in part, in the Ph.D. thesis of Angela Michelle Getz, in the

Department of Neuroscience, Faculty of Graduate Studies, University of Calgary.

The thesis will be published in the Library and Archives Canada (through the University of

Calgary). To allow this, Angela Michelle Getz has signed a Theses Non-Exclusive License that authorizes Library and Archives Canada to reproduce, communicate to the public on the Internet, loan, distribute or sell copies of the above listed thesis, among other things. By signing this document, I, Fenglian Xu, indicate that I have read and understand this agreement and I allow the listed work to be published by Library and Archives Canada.

(Signature removed)

______

Dr. Fenglian Xu

September 15, 2016

______

Date

276

Dr. Nichole Flynn (University of Calgary) is a co-author on the following manuscript:

This manuscript has been reproduced, in part, in the Ph.D. thesis of Angela Michelle Getz, in the

Department of Neuroscience, Faculty of Graduate Studies, University of Calgary.

The thesis will be published in the Library and Archives Canada (through the University of

Calgary). To allow this, Angela Michelle Getz has signed a Theses Non-Exclusive License that authorizes Library and Archives Canada to reproduce, communicate to the public on the Internet, loan, distribute or sell copies of the above listed thesis, among other things. By signing this document, I, Nichole Flynn, indicate that I have read and understand this agreement and I allow the listed work to be published by Library and Archives Canada.

(Signature removed)

______

Dr. Nichole Flynn

September 14, 2016

______

Date

277

Wali Zaidi (University of Calgary) is a co-author on the following manuscript:

This manuscript has been reproduced, in part, in the Ph.D. thesis of Angela Michelle Getz, in the

Department of Neuroscience, Faculty of Graduate Studies, University of Calgary.

The thesis will be published in the Library and Archives Canada (through the University of

Calgary). To allow this, Angela Michelle Getz has signed a Theses Non-Exclusive License that

authorizes Library and Archives Canada to reproduce, communicate to the public on the Internet,

loan, distribute or sell copies of the above listed thesis, among other things. By signing this

document, I, Wali Zaidi, indicate that I have read and understand this agreement and I allow the listed work to be published by Library and Archives Canada.

(Signature removed)

______

Wali Zaidi

November 12, 2016

______

Date

278

Dr. Naweed Syed (University of Calgary) is a co-author on the following manuscript:

This manuscript has been reproduced, in part, in the Ph.D. thesis of Angela Michelle Getz, in the

Department of Neuroscience, Faculty of Graduate Studies, University of Calgary.

The thesis will be published in the Library and Archives Canada (through the University of

Calgary). To allow this, Angela Michelle Getz has signed a Theses Non-Exclusive License that

authorizes Library and Archives Canada to reproduce, communicate to the public on the Internet,

loan, distribute or sell copies of the above listed thesis, among other things. By signing this

document, I, Naweed Syed, indicate that I have read and understand this agreement and I allow the listed work to be published by Library and Archives Canada.

(Signature removed)

______

Dr. Naweed Syed

September 14, 2016

______

Date

279

Chapter Three: Tumor suppressor menin is required for subunit-specific nAChR α5 transcription and nAChR-dependent presynaptic facilitation in cultured mouse hippocampal neurons (submitted manuscript)

Collaborators: Angela M. Getz, Fenglian Xu, Frank Visser, Pia C. Christensen, Roger Persson,

Naweed I. Syed.

Contributions: Experimental design: AG, FV, NS. Manuscript and figure preparation: AG. Cell culture and tissue preparation: AG, FX (embryonic hippocampi dissections, primary cell culture - partial), PC (5xFAD hippocampi dissections). Electrophysiology: AG, FX (patch clamp recordings - partial). Molecular biology: AG. ICC: AG. Imaging: AG, RP (super-resolution imaging).

Funding: This work was supported by a Canadian Institute of Health Research (CIHR) grant to

NS. FX was funded by a Natural Sciences and Engineering Research Council of Canada (NSERC) grant. AG was funded by Alberta Innovates – Health Solutions (AIHS) and NSERC studentships.

This work was supported by the Molecular Biology Core Facility and the Regeneration Unit in

Neurobiology (RUN) Advanced Optical Microscopy Imaging Core Facility of the Hotchkiss Brain

Institute at the University of Calgary (Calgary, Canada). Super resolution imaging was performed by BioAxial (Paris, France), which commercializes CoDiM technology. RP is an employee of

BioAxial.

Competing interests: None.

280

Chapter Four: A novel mechanism of synaptic specificity: synaptophysin inhibits

presynaptic secretory machinery to regulate the release of peptide neurotransmitters

(manuscript in preparation)

Collaborators: Angela M. Getz, Tara A. Janes, Frank Visser, Wali Zaidi, Amar Deshwar, Jean

Kawasoe, Naweed I. Syed.

Contributions: Experimental design: AG, NS. Manuscript and figure preparation: AG. Cell culture

and tissue preparation: WZ (synaptic and single cell culture), AG, TJ (intact CNS preparations),

AD (tissue dissections and cryostat sectioning - partial), JK (cryostat sectioning - partial).

Electrophysiology: AG, TJ (intact CNS recordings). Molecular biology: AG. ICC: AG. IHC &

ISH: AD, JK. Imaging: AG, TJ (intact CNS LY images).

Funding: This work was supported by a Canadian Institutes of Health Research (CIHR) grant to

NS. AG was funded by Alberta Innovates – Health Solutions (AIHS) and Natural Sciences and

Engineering Research Council of Canada (NSERC) studentships. AD was funded by an AIHS summer studentship. Molecular biology and imaging experiments were supported by the

Molecular Biology Core Facility and the Regeneration Unit in Neurobiology Advanced Optical

Microscopy Imaging Core Facility of the Hotchkiss Brain Institute at the University of Calgary.

Competing interests: None.

281

References

1 Lisman, J. E., Raghavachari, S. & Tsien, R. W. The sequence of events that underlie quantal transmission at central glutamatergic synapses. Nature reviews. Neuroscience 8, 597-609, doi:10.1038/nrn2191 (2007). 2 Ribrault, C., Sekimoto, K. & Triller, A. From the stochasticity of molecular processes to the variability of synaptic transmission. Nature reviews. Neuroscience 12, 375-387, doi:10.1038/nrn3025 (2011). 3 Kupfermann, I. Functional studies of cotransmission. Physiological reviews 71, 683-732 (1991). 4 Carta, M. et al. Membrane lipids tune synaptic transmission by direct modulation of presynaptic potassium channels. Neuron 81, 787-799, doi:10.1016/j.neuron.2013.12.028 (2014). 5 Gotti, C. et al. Heterogeneity and complexity of native brain nicotinic receptors. Biochemical pharmacology 74, 1102-1111, doi:10.1016/j.bcp.2007.05.023 (2007). 6 Wang, F. et al. Assembly of human neuronal nicotinic receptor alpha5 subunits with alpha3, beta2, and beta4 subunits. The Journal of biological chemistry 271, 17656-17665 (1996). 7 Ramirez-Latorre, J. et al. Functional contributions of alpha5 subunit to neuronal acetylcholine receptor channels. Nature 380, 347-351, doi:10.1038/380347a0 (1996). 8 Vincler, M. A. & Eisenach, J. C. Knock down of the alpha 5 nicotinic acetylcholine receptor in spinal nerve-ligated rats alleviates mechanical allodynia. Pharmacology, biochemistry, and behavior 80, 135-143, doi:10.1016/j.pbb.2004.10.011 (2005). 9 Cottrell, G. A. & Davies, N. W. Multiple receptor sites for a molluscan peptide (FMRFamide) and related peptides of Helix. The Journal of physiology 382, 51-68 (1987). 10 Brussaard, A. B., Kits, K. S. & ter Maat, A. One receptor type mediates two independent effects of FMRFa on neurosecretory cells of Lymnaea. Peptides 10, 289-297 (1989). 11 Sudhof, T. C. The synaptic vesicle cycle. Annual review of neuroscience 27, 509-547, doi:10.1146/annurev.neuro.26.041002.131412 (2004). 12 Fernandez-Chacon, R. et al. Synaptotagmin I functions as a calcium regulator of release probability. Nature 410, 41-49, doi:10.1038/35065004 (2001). 13 Sollner, T. et al. SNAP receptors implicated in vesicle targeting and fusion. Nature 362, 318-324, doi:10.1038/362318a0 (1993). 14 Bahler, M. & Greengard, P. Synapsin I bundles F-actin in a phosphorylation-dependent manner. Nature 326, 704-707, doi:10.1038/326704a0 (1987). 15 Jackman, S. L., Turecek, J., Belinsky, J. E. & Regehr, W. G. The calcium sensor synaptotagmin 7 is required for synaptic facilitation. Nature 529, 88-91, doi:10.1038/nature16507 (2016). 16 McMahon, H. T. et al. Synaptophysin, a major synaptic vesicle protein, is not essential for neurotransmitter release. Proceedings of the National Academy of Sciences of the United States of America 93, 4760-4764 (1996). 17 Thomas, L. et al. Identification of synaptophysin as a hexameric channel protein of the synaptic vesicle membrane. Science 242, 1050-1053 (1988).

282

18 Edelmann, L., Hanson, P. I., Chapman, E. R. & Jahn, R. Synaptobrevin binding to synaptophysin: a potential mechanism for controlling the exocytotic fusion machine. The EMBO journal 14, 224-231 (1995). 19 Janz, R. et al. Essential roles in synaptic plasticity for synaptogyrin I and synaptophysin I. Neuron 24, 687-700 (1999). 20 Navone, F., Greengard, P. & De Camilli, P. Synapsin I in nerve terminals: selective association with small synaptic vesicles. Science 226, 1209-1211 (1984). 21 Navone, F. et al. Protein p38: an integral membrane protein specific for small vesicles of neurons and neuroendocrine cells. The Journal of cell biology 103, 2511-2527 (1986). 22 Mutch, S. A. et al. Protein quantification at the single vesicle level reveals that a subset of synaptic vesicle proteins are trafficked with high precision. The Journal of neuroscience : the official journal of the Society for Neuroscience 31, 1461-1470, doi:10.1523/JNEUROSCI.3805-10.2011 (2011). 23 Salio, C., Lossi, L., Ferrini, F. & Merighi, A. Neuropeptides as synaptic transmitters. Cell and tissue research 326, 583-598, doi:10.1007/s00441-006-0268-3 (2006). 24 Whim, M. D. & Lloyd, P. E. Frequency-dependent release of peptide cotransmitters from identified cholinergic motor neurons in Aplysia. Proceedings of the National Academy of Sciences of the United States of America 86, 9034-9038 (1989). 25 Borbely, E., Scheich, B. & Helyes, Z. Neuropeptides in learning and memory. Neuropeptides 47, 439-450, doi:10.1016/j.npep.2013.10.012 (2013). 26 Budnik, V. et al. Regulation of synapse structure and function by the Drosophila tumor suppressor gene dlg. Neuron 17, 627-640 (1996). 27 El-Husseini, A. E., Schnell, E., Chetkovich, D. M., Nicoll, R. A. & Bredt, D. S. PSD-95 involvement in maturation of excitatory synapses. Science 290, 1364-1368 (2000). 28 Kornau, H. C., Schenker, L. T., Kennedy, M. B. & Seeburg, P. H. Domain interaction between NMDA receptor subunits and the postsynaptic density protein PSD-95. Science 269, 1737-1740 (1995). 29 Dong, H. et al. GRIP: a synaptic PDZ domain-containing protein that interacts with AMPA receptors. Nature 386, 279-284, doi:10.1038/386279a0 (1997). 30 Xia, J., Zhang, X., Staudinger, J. & Huganir, R. L. Clustering of AMPA receptors by the synaptic PDZ domain-containing protein PICK1. Neuron 22, 179-187 (1999). 31 Gautam, M. et al. Failure of postsynaptic specialization to develop at neuromuscular junctions of rapsyn-deficient mice. Nature 377, 232-236, doi:10.1038/377232a0 (1995). 32 Ramarao, M. K. & Cohen, J. B. Mechanism of nicotinic acetylcholine receptor cluster formation by rapsyn. Proceedings of the National Academy of Sciences of the United States of America 95, 4007-4012 (1998). 33 Feng, G., Steinbach, J. H. & Sanes, J. R. Rapsyn clusters neuronal acetylcholine receptors but is inessential for formation of an interneuronal cholinergic synapse. The Journal of neuroscience : the official journal of the Society for Neuroscience 18, 4166-4176 (1998). 34 Yakel, J. L. Nicotinic ACh receptors in the hippocampal circuit; functional expression and role in synaptic plasticity. The Journal of physiology 592, 4147-4153, doi:10.1113/jphysiol.2014.273896 (2014). 35 Young, S. H. & Poo, M. M. Spontaneous release of transmitter from growth cones of embryonic neurones. Nature 305, 634-637 (1983).

283

36 Hume, R. I., Role, L. W. & Fischbach, G. D. Acetylcholine release from growth cones detected with patches of acetylcholine receptor-rich membranes. Nature 305, 632-634 (1983). 37 Chow, I. & Poo, M. M. Release of acetylcholine from embryonic neurons upon contact with muscle cell. The Journal of neuroscience : the official journal of the Society for Neuroscience 5, 1076-1082 (1985). 38 Haydon, P. G., McCobb, D. P. & Kater, S. B. Serotonin selectively inhibits growth cone motility and synaptogenesis of specific identified neurons. Science 226, 561-564 (1984). 39 Fischer, M., Kaech, S., Wagner, U., Brinkhaus, H. & Matus, A. Glutamate receptors regulate actin-based plasticity in dendritic spines. Nature neuroscience 3, 887-894, doi:10.1038/78791 (2000). 40 Haydon, P. G. & Drapeau, P. From contact to connection: early events during synaptogenesis. Trends in neurosciences 18, 196-201 (1995). 41 Luk, C. C., Schmold, N. M., Lee, T. K. & Syed, N. I. A novel approach reveals temporal patterns of synaptogenesis between the isolated growth cones of Lymnaea neurons. The European journal of neuroscience 32, 1442-1451, doi:10.1111/j.1460-9568.2010.07428.x (2010). 42 Kurschner, C., Mermelstein, P. G., Holden, W. T. & Surmeier, D. J. CIPP, a novel multivalent PDZ domain protein, selectively interacts with Kir4.0 family members, NMDA receptor subunits, neurexins, and neuroligins. Molecular and cellular neurosciences 11, 161-172, doi:10.1006/mcne.1998.0679 (1998). 43 Biederer, T. & Sudhof, T. C. Mints as adaptors. Direct binding to neurexins and recruitment of munc18. The Journal of biological chemistry 275, 39803-39806, doi:10.1074/jbc.C000656200 (2000). 44 Scheiffele, P., Fan, J., Choih, J., Fetter, R. & Serafini, T. Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell 101, 657- 669 (2000). 45 Graf, E. R., Zhang, X., Jin, S. X., Linhoff, M. W. & Craig, A. M. Neurexins induce differentiation of GABA and glutamate postsynaptic specializations via neuroligins. Cell 119, 1013-1026, doi:10.1016/j.cell.2004.11.035 (2004). 46 Chubykin, A. A. et al. Activity-dependent validation of excitatory versus inhibitory synapses by neuroligin-1 versus neuroligin-2. Neuron 54, 919-931, doi:10.1016/j.neuron.2007.05.029 (2007). 47 Dean, C. et al. Neurexin mediates the assembly of presynaptic terminals. Nature neuroscience 6, 708-716, doi:10.1038/nn1074 (2003). 48 Zhai, R. G. et al. Assembling the presynaptic active zone: a characterization of an active one precursor vesicle. Neuron 29, 131-143 (2001). 49 Shapira, M. et al. Unitary assembly of presynaptic active zones from Piccolo-Bassoon transport vesicles. Neuron 38, 237-252 (2003). 50 Gerrow, K. et al. A preformed complex of postsynaptic proteins is involved in excitatory synapse development. Neuron 49, 547-562, doi:10.1016/j.neuron.2006.01.015 (2006). 51 Hayashi, M. K. et al. The postsynaptic density proteins Homer and Shank form a polymeric network structure. Cell 137, 159-171, doi:10.1016/j.cell.2009.01.050 (2009). 52 Bresler, T. et al. Postsynaptic density assembly is fundamentally different from presynaptic active zone assembly. The Journal of neuroscience : the official journal of

284

the Society for Neuroscience 24, 1507-1520, doi:10.1523/JNEUROSCI.3819-03.2004 (2004). 53 Bresler, T. et al. The dynamics of SAP90/PSD-95 recruitment to new synaptic junctions. Molecular and cellular neurosciences 18, 149-167, doi:10.1006/mcne.2001.1012 (2001). 54 Friedman, H. V., Bresler, T., Garner, C. C. & Ziv, N. E. Assembly of new individual excitatory synapses: time course and temporal order of synaptic molecule recruitment. Neuron 27, 57-69 (2000). 55 Lin, S., Landmann, L., Ruegg, M. A. & Brenner, H. R. The role of nerve- versus muscle- derived factors in mammalian neuromuscular junction formation. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 3333-3340, doi:10.1523/JNEUROSCI.5590-07.2008 (2008). 56 Goda, Y. & Davis, G. W. Mechanisms of synapse assembly and disassembly. Neuron 40, 243-264 (2003). 57 Verhage, M. et al. Synaptic assembly of the brain in the absence of neurotransmitter secretion. Science 287, 864-869 (2000). 58 Katz, L. C. & Shatz, C. J. Synaptic activity and the construction of cortical circuits. Science 274, 1133-1138 (1996). 59 Wong, R. O. Retinal waves and visual system development. Annual review of neuroscience 22, 29-47, doi:10.1146/annurev.neuro.22.1.29 (1999). 60 Hanse, E., Seth, H. & Riebe, I. AMPA-silent synapses in brain development and pathology. Nature reviews. Neuroscience 14, 839-850, doi:10.1038/nrn3642 (2013). 61 Hannan, A. J. et al. PLC-beta1, activated via mGluRs, mediates activity-dependent differentiation in cerebral cortex. Nature neuroscience 4, 282-288, doi:10.1038/85132 (2001). 62 Wijetunge, L. S., Till, S. M., Gillingwater, T. H., Ingham, C. A. & Kind, P. C. mGluR5 regulates glutamate-dependent development of the mouse somatosensory cortex. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 13028- 13037, doi:10.1523/JNEUROSCI.2600-08.2008 (2008). 63 Gomperts, S. N., Carroll, R., Malenka, R. C. & Nicoll, R. A. Distinct roles for ionotropic and metabotropic glutamate receptors in the maturation of excitatory synapses. The Journal of neuroscience : the official journal of the Society for Neuroscience 20, 2229- 2237 (2000). 64 Meier, J., Juttner, R., Kirischuk, S. & Grantyn, R. Synaptic anchoring of glycine receptors in developing collicular neurons under control of metabotropic glutamate receptor activity. Molecular and cellular neurosciences 21, 324-340 (2002). 65 Shakiryanova, D., Tully, A., Hewes, R. S., Deitcher, D. L. & Levitan, E. S. Activity- dependent liberation of synaptic neuropeptide vesicles. Nature neuroscience 8, 173-178, doi:10.1038/nn1377 (2005). 66 Shakiryanova, D., Tully, A. & Levitan, E. S. Activity-dependent synaptic capture of transiting peptidergic vesicles. Nature neuroscience 9, 896-900, doi:10.1038/nn1719 (2006). 67 Knox, R. J., Quattrocki, E. A., Connor, J. A. & Kaczmarek, L. K. Recruitment of Ca2+ channels by protein kinase C during rapid formation of putative neuropeptide release sites in isolated Aplysia neurons. Neuron 8, 883-889 (1992).

285

68 Voronezhskaya, E. E. & Elekes, K. Expression of FMRFamide gene encoded peptides by identified neurons in embryos and juveniles of the pulmonate snail Lymnaea stagnalis. Cell and tissue research 314, 297-313, doi:10.1007/s00441-003-0800-7 (2003). 69 Voronezhskaya, E. E. & Elekes, K. Transient and sustained expression of FMRFamide- like immunoreactivity in the developing nervous system of Lymnaea stagnalis (Mollusca, Pulmonata). Cellular and molecular neurobiology 16, 661-676 (1996). 70 Blum, R. & Konnerth, A. Neurotrophin-mediated rapid signaling in the central nervous system: mechanisms and functions. Physiology 20, 70-78, doi:10.1152/physiol.00042.2004 (2005). 71 Park, H. & Poo, M. M. Neurotrophin regulation of neural circuit development and function. Nature reviews. Neuroscience 14, 7-23, doi:10.1038/nrn3379 (2013). 72 Lohof, A. M., Ip, N. Y. & Poo, M. M. Potentiation of developing neuromuscular synapses by the neurotrophins NT-3 and BDNF. Nature 363, 350-353, doi:10.1038/363350a0 (1993). 73 Jovanovic, J. N., Czernik, A. J., Fienberg, A. A., Greengard, P. & Sihra, T. S. Synapsins as mediators of BDNF-enhanced neurotransmitter release. Nature neuroscience 3, 323- 329, doi:10.1038/73888 (2000). 74 Shen, W. et al. Activity-induced rapid synaptic maturation mediated by presynaptic cdc42 signaling. Neuron 50, 401-414, doi:10.1016/j.neuron.2006.03.017 (2006). 75 Flores-Otero, J., Xue, H. Z. & Davis, R. L. Reciprocal regulation of presynaptic and postsynaptic proteins in bipolar spiral ganglion neurons by neurotrophins. The Journal of neuroscience : the official journal of the Society for Neuroscience 27, 14023-14034, doi:10.1523/JNEUROSCI.3219-07.2007 (2007). 76 Sanes, J. R. & Lichtman, J. W. Development of the vertebrate neuromuscular junction. Annual review of neuroscience 22, 389-442, doi:10.1146/annurev.neuro.22.1.389 (1999). 77 Ghazanfari, N. et al. Muscle specific kinase: organiser of synaptic membrane domains. The international journal of biochemistry & cell biology 43, 295-298, doi:10.1016/j.biocel.2010.10.008 (2011). 78 Borges, L. S. et al. Identification of a motif in the acetylcholine receptor beta subunit whose phosphorylation regulates rapsyn association and postsynaptic receptor localization. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 11468-11476, doi:10.1523/JNEUROSCI.2508-08.2008 (2008). 79 Je, H. S. et al. Role of pro-brain-derived neurotrophic factor (proBDNF) to mature BDNF conversion in activity-dependent competition at developing neuromuscular synapses. Proceedings of the National Academy of Sciences of the United States of America 109, 15924-15929, doi:10.1073/pnas.1207767109 (2012). 80 Gianfranceschi, L. et al. Visual cortex is rescued from the effects of dark rearing by overexpression of BDNF. Proceedings of the National Academy of Sciences of the United States of America 100, 12486-12491, doi:10.1073/pnas.1934836100 (2003). 81 Kafitz, K. W., Rose, C. R., Thoenen, H. & Konnerth, A. Neurotrophin-evoked rapid excitation through TrkB receptors. Nature 401, 918-921, doi:10.1038/44847 (1999). 82 Blum, R., Kafitz, K. W. & Konnerth, A. Neurotrophin-evoked depolarization requires the sodium channel Na(V)1.9. Nature 419, 687-693, doi:10.1038/nature01085 (2002).

286

83 Greer, P. L. & Greenberg, M. E. From synapse to nucleus: calcium-dependent gene transcription in the control of synapse development and function. Neuron 59, 846-860, doi:10.1016/j.neuron.2008.09.002 (2008). 84 Bading, H. Nuclear calcium signalling in the regulation of brain function. Nature reviews. Neuroscience 14, 593-608, doi:10.1038/nrn3531 (2013). 85 Greenberg, M. E., Ziff, E. B. & Greene, L. A. Stimulation of neuronal acetylcholine receptors induces rapid gene transcription. Science 234, 80-83 (1986). 86 Velho, T. A. & Mello, C. V. Synapsins are late activity-induced genes regulated by birdsong. The Journal of neuroscience : the official journal of the Society for Neuroscience 28, 11871-11882, doi:10.1523/JNEUROSCI.2307-08.2008 (2008). 87 Jo, S. A., Zhu, X., Marchionni, M. A. & Burden, S. J. Neuregulins are concentrated at nerve-muscle synapses and activate ACh-receptor gene expression. Nature 373, 158-161, doi:10.1038/373158a0 (1995). 88 Sandrock, A. W., Jr. et al. Maintenance of acetylcholine receptor number by neuregulins at the neuromuscular junction in vivo. Science 276, 599-603 (1997). 89 Schaeffer, L., Duclert, N., Huchet-Dymanus, M. & Changeux, J. P. Implication of a multisubunit Ets-related transcription factor in synaptic expression of the nicotinic acetylcholine receptor. The EMBO journal 17, 3078-3090, doi:10.1093/emboj/17.11.3078 (1998). 90 Yang, X., Kuo, Y., Devay, P., Yu, C. & Role, L. A cysteine-rich isoform of neuregulin controls the level of expression of neuronal nicotinic receptor channels during synaptogenesis. Neuron 20, 255-270 (1998). 91 Kim, H. G., Cho, S. M., Lee, C. K. & Jeong, S. W. Neuregulin 1 as an endogenous regulator of nicotinic acetylcholine receptors in adult major pelvic ganglion neurons. Biochemical and biophysical research communications 463, 632-637, doi:10.1016/j.bbrc.2015.05.113 (2015). 92 Liu, Y., Ford, B., Mann, M. A. & Fischbach, G. D. Neuregulins increase alpha7 nicotinic acetylcholine receptors and enhance excitatory synaptic transmission in GABAergic interneurons of the hippocampus. The Journal of neuroscience : the official journal of the Society for Neuroscience 21, 5660-5669 (2001). 93 Ozaki, M., Sasner, M., Yano, R., Lu, H. S. & Buonanno, A. Neuregulin-beta induces expression of an NMDA-receptor subunit. Nature 390, 691-694, doi:10.1038/37795 (1997). 94 Caldeira, M. V. et al. Brain-derived neurotrophic factor regulates the expression and synaptic delivery of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor subunits in hippocampal neurons. The Journal of biological chemistry 282, 12619-12628, doi:10.1074/jbc.M700607200 (2007). 95 Dash, P. K., Hochner, B. & Kandel, E. R. Injection of the cAMP-responsive element into the nucleus of Aplysia sensory neurons blocks long-term facilitation. Nature 345, 718- 721, doi:10.1038/345718a0 (1990). 96 Sheng, M., Thompson, M. A. & Greenberg, M. E. CREB: a Ca(2+)-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science 252, 1427- 1430 (1991).

287

97 Xing, J., Ginty, D. D. & Greenberg, M. E. Coupling of the RAS-MAPK pathway to gene activation by RSK2, a growth factor-regulated CREB kinase. Science 273, 959-963 (1996). 98 Dash, P. K., Karl, K. A., Colicos, M. A., Prywes, R. & Kandel, E. R. cAMP response element-binding protein is activated by Ca2+/calmodulin- as well as cAMP-dependent protein kinase. Proceedings of the National Academy of Sciences of the United States of America 88, 5061-5065 (1991). 99 van Kesteren, R. E. et al. Synapse formation between central neurons requires postsynaptic expression of the MEN1 tumor suppressor gene. The Journal of neuroscience : the official journal of the Society for Neuroscience 21, RC161 (2001). 100 van Kesteren, R. E. et al. Postsynaptic expression of an epidermal growth factor receptor regulates cholinergic synapse formation between identified molluscan neurons. The European journal of neuroscience 27, 2043-2056, doi:10.1111/j.1460-9568.2008.06189.x (2008). 101 Xu, F., Hennessy, D. A., Lee, T. K. & Syed, N. I. Trophic factor-induced intracellular calcium oscillations are required for the expression of postsynaptic acetylcholine receptors during synapse formation between Lymnaea neurons. The Journal of neuroscience : the official journal of the Society for Neuroscience 29, 2167-2176, doi:10.1523/JNEUROSCI.4682-08.2009 (2009). 102 Flynn, N., Getz, A., Visser, F., Janes, T. A. & Syed, N. I. Menin: A Tumor Suppressor That Mediates Postsynaptic Receptor Expression and Synaptogenesis between Central Neurons of Lymnaea stagnalis. PloS one 9, e111103, doi:10.1371/journal.pone.0111103 (2014). 103 Woolf, C. J., Shortland, P. & Coggeshall, R. E. Peripheral nerve injury triggers central sprouting of myelinated afferents. Nature 355, 75-78, doi:10.1038/355075a0 (1992). 104 Shen, H., Chung, J. M. & Chung, K. Expression of neurotrophin mRNAs in the dorsal root ganglion after spinal nerve injury. Brain research. Molecular brain research 64, 186-192 (1999). 105 Zhang, X., Chen, G., Xue, Q. & Yu, B. Early changes of beta-Catenins and Menins in spinal cord dorsal horn after peripheral nerve injury. Cellular and molecular neurobiology 30, 885-890, doi:10.1007/s10571-010-9517-9 (2010). 106 Xu, S. et al. Tumor suppressor menin mediates peripheral nerve injury-induced neuropathic pain through potentiating synaptic plasticity. Neuroscience 223, 473-485, doi:10.1016/j.neuroscience.2012.07.036 (2012). 107 Shen, X. et al. Menin regulates spinal glutamate-GABA balance through GAD65 contributing to neuropathic pain. Pharmacological reports : PR 66, 49-55, doi:10.1016/j.pharep.2013.06.005 (2014). 108 Santini, E. & Klann, E. Reciprocal signaling between translational control pathways and synaptic proteins in autism spectrum disorders. Science signaling 7, re10, doi:10.1126/scisignal.2005832 (2014). 109 Knafo, S. et al. PTEN recruitment controls synaptic and cognitive function in Alzheimer's models. Nature neuroscience, doi:10.1038/nn.4225 (2016). 110 Suberbielle, E. et al. DNA repair factor BRCA1 depletion occurs in Alzheimer brains and impairs cognitive function in mice. Nature communications 6, 8897, doi:10.1038/ncomms9897 (2015).

288

111 Matkar, S., Thiel, A. & Hua, X. Menin: a scaffold protein that controls gene expression and cell signaling. Trends in biochemical sciences 38, 394-402, doi:10.1016/j.tibs.2013.05.005 (2013). 112 Zipursky, S. L. & Sanes, J. R. Chemoaffinity revisited: dscams, protocadherins, and neural circuit assembly. Cell 143, 343-353, doi:10.1016/j.cell.2010.10.009 (2010). 113 Williams, M. E., de Wit, J. & Ghosh, A. Molecular mechanisms of synaptic specificity in developing neural circuits. Neuron 68, 9-18, doi:10.1016/j.neuron.2010.09.007 (2010). 114 Williams, M. E. et al. Cadherin-9 regulates synapse-specific differentiation in the developing hippocampus. Neuron 71, 640-655, doi:10.1016/j.neuron.2011.06.019 (2011). 115 Craig, A. M. & Boudin, H. Molecular heterogeneity of central synapses: afferent and target regulation. Nature neuroscience 4, 569-578, doi:10.1038/88388 (2001). 116 Terauchi, A. et al. Distinct FGFs promote differentiation of excitatory and inhibitory synapses. Nature 465, 783-787, doi:10.1038/nature09041 (2010). 117 Mayford, M., Barzilai, A., Keller, F., Schacher, S. & Kandel, E. R. Modulation of an NCAM-related adhesion molecule with long-term synaptic plasticity in Aplysia. Science 256, 638-644 (1992). 118 Zhu, H., Wu, F. & Schacher, S. Changes in expression and distribution of Aplysia cell adhesion molecules can influence synapse formation and elimination in vitro. The Journal of neuroscience : the official journal of the Society for Neuroscience 15, 4173- 4183 (1995). 119 Sweatt, J. D. et al. FMRFamide reverses protein phosphorylation produced by 5-HT and cAMP in Aplysia sensory neurons. Nature 342, 275-278, doi:10.1038/342275a0 (1989). 120 Guan, Z. et al. Integration of long-term-memory-related synaptic plasticity involves bidirectional regulation of gene expression and chromatin structure. Cell 111, 483-493 (2002). 121 Martin, K. C. et al. Synapse-specific, long-term facilitation of aplysia sensory to motor synapses: a function for local protein synthesis in memory storage. Cell 91, 927-938 (1997). 122 Casadio, A. et al. A transient, neuron-wide form of CREB-mediated long-term facilitation can be stabilized at specific synapses by local protein synthesis. Cell 99, 221- 237 (1999). 123 Lyles, V., Zhao, Y. & Martin, K. C. Synapse formation and mRNA localization in cultured Aplysia neurons. Neuron 49, 349-356, doi:10.1016/j.neuron.2005.12.029 (2006). 124 Wang, D. O. et al. Synapse- and stimulus-specific local translation during long-term neuronal plasticity. Science 324, 1536-1540, doi:10.1126/science.1173205 (2009). 125 Puthanveettil, S. V. et al. A new component in synaptic plasticity: upregulation of kinesin in the neurons of the gill-withdrawal reflex. Cell 135, 960-973, doi:10.1016/j.cell.2008.11.003 (2008). 126 Rosenmund, C., Clements, J. D. & Westbrook, G. L. Nonuniform probability of glutamate release at a hippocampal synapse. Science 262, 754-757 (1993). 127 Samano, C., Cifuentes, F. & Morales, M. A. Neurotransmitter segregation: functional and plastic implications. Progress in neurobiology 97, 277-287, doi:10.1016/j.pneurobio.2012.04.004 (2012). 128 Maccaferri, G., Toth, K. & McBain, C. J. Target-specific expression of presynaptic mossy fiber plasticity. Science 279, 1368-1370 (1998).

289

129 Davis, G. W. & Murphey, R. K. A role for postsynaptic neurons in determining presynaptic release properties in the cricket CNS: evidence for retrograde control of facilitation. The Journal of neuroscience : the official journal of the Society for Neuroscience 13, 3827-3838 (1993). 130 Katz, P. S., Kirk, M. D. & Govind, C. K. Facilitation and depression at different branches of the same motor axon: evidence for presynaptic differences in release. The Journal of neuroscience : the official journal of the Society for Neuroscience 13, 3075-3089 (1993). 131 Laurent, G. & Sivaramakrishnan, A. Single local interneurons in the locust make central synapses with different properties of transmitter release on distinct postsynaptic neurons. The Journal of neuroscience : the official journal of the Society for Neuroscience 12, 2370-2380 (1992). 132 Reyes, A. et al. Target-cell-specific facilitation and depression in neocortical circuits. Nature neuroscience 1, 279-285, doi:10.1038/1092 (1998). 133 Muller, K. J. & Nicholls, J. G. Different properties of synapses between a single sensory neurone and two different motor cells in the leech C.N.S. The Journal of physiology 238, 357-369 (1974). 134 Pelkey, K. A., Lavezzari, G., Racca, C., Roche, K. W. & McBain, C. J. mGluR7 is a metaplastic switch controlling bidirectional plasticity of feedforward inhibition. Neuron 46, 89-102, doi:10.1016/j.neuron.2005.02.011 (2005). 135 Shigemoto, R. et al. Target-cell-specific concentration of a metabotropic glutamate receptor in the presynaptic active zone. Nature 381, 523-525, doi:10.1038/381523a0 (1996). 136 Suh, Y. H. et al. Corequirement of PICK1 binding and PKC phosphorylation for stable surface expression of the metabotropic glutamate receptor mGluR7. Neuron 58, 736-748, doi:10.1016/j.neuron.2008.03.028 (2008). 137 Schacher, S., Rayport, S. G. & Ambron, R. T. Giant Aplysia neuron R2 reliably forms strong chemical connections in vitro. The Journal of neuroscience : the official journal of the Society for Neuroscience 5, 2851-2856 (1985). 138 Whim, M. D., Niemann, H. & Kaczmarek, L. K. The secretion of classical and peptide cotransmitters from a single presynaptic neuron involves a synaptobrevin-like molecule. The Journal of neuroscience : the official journal of the Society for Neuroscience 17, 2338-2347 (1997). 139 Blitz, D. M. & Nusbaum, M. P. Distinct functions for cotransmitters mediating motor pattern selection. The Journal of neuroscience : the official journal of the Society for Neuroscience 19, 6774-6783 (1999). 140 Vullhorst, D. et al. Selective expression of ErbB4 in interneurons, but not pyramidal cells, of the rodent hippocampus. The Journal of neuroscience : the official journal of the Society for Neuroscience 29, 12255-12264, doi:10.1523/JNEUROSCI.2454-09.2009 (2009). 141 Sun, Q. Q. et al. Differential metabotropic glutamate receptor expression and modulation in two neocortical inhibitory networks. Journal of neurophysiology 101, 2679-2692, doi:10.1152/jn.90566.2008 (2009). 142 Schmold, N. & Syed, N. I. Molluscan neurons in culture: shedding light on synapse formation and plasticity. Journal of molecular histology 43, 383-399, doi:10.1007/s10735-012-9398-y (2012).

290

143 McCroskery, S., Bailey, A., Lin, L. & Daniels, M. P. Transmembrane agrin regulates dendritic filopodia and synapse formation in mature hippocampal neuron cultures. Neuroscience 163, 168-179, doi:10.1016/j.neuroscience.2009.06.012 (2009). 144 Syed, N. I., Bulloch, A. G. & Lukowiak, K. In vitro reconstruction of the respiratory central pattern generator of the mollusk Lymnaea. Science 250, 282-285 (1990). 145 Syed, N. I., Ridgway, R. L., Lukowiak, K. & Bulloch, A. G. Transplantation and functional integration of an identified respiratory interneuron in Lymnaea stagnalis. Neuron 8, 767-774 (1992). 146 Crossley, M., Staras, K. & Kemenes, G. A two-neuron system for adaptive goal-directed decision-making in Lymnaea. Nature communications 7, 11793, doi:10.1038/ncomms11793 (2016). 147 Kaczmarek, L. K., Finbow, M., Revel, J. P. & Strumwasser, F. The morphology and coupling of Aplysia bag cells within the abdominal ganglion and in cell culture. Journal of neurobiology 10, 535-550, doi:10.1002/neu.480100604 (1979). 148 Ready, D. F. & Nicholls, J. Identified neurones isolated from leech CNS make selective connections in culture. Nature 281, 67-69 (1979). 149 Bulloch, A. G. & Syed, N. I. Reconstruction of neuronal networks in culture. Trends in neurosciences 15, 422-427 (1992). 150 Feng, Z. P. et al. Transcriptome analysis of the central nervous system of the mollusc Lymnaea stagnalis. BMC genomics 10, 451, doi:10.1186/1471-2164-10-451 (2009). 151 Ridgway, R. L., Syed, N. I., Lukowiak, K. & Bulloch, A. G. Nerve growth factor (NGF) induces sprouting of specific neurons of the snail, Lymnaea stagnalis. Journal of neurobiology 22, 377-390, doi:10.1002/neu.480220406 (1991). 152 Feng, Z. P., Klumperman, J., Lukowiak, K. & Syed, N. I. In vitro synaptogenesis between the somata of identified Lymnaea neurons requires protein synthesis but not extrinsic growth factors or substrate adhesion molecules. The Journal of neuroscience : the official journal of the Society for Neuroscience 17, 7839-7849 (1997). 153 Liu, Y. & Nicholls, J. Steps in the development of chemical and electrical synapses by pairs of identified leech neurons in culture. Proceedings of the Royal Society of London. Series B, Biological sciences 236, 253-268 (1989). 154 Ohnuma, K., Whim, M. D., Fetter, R. D., Kaczmarek, L. K. & Zucker, R. S. Presynaptic target of Ca2+ action on neuropeptide and acetylcholine release in Aplysia californica. The Journal of physiology 535, 647-662 (2001). 155 Feng, Z. P. et al. Development of Ca2+ hotspots between Lymnaea neurons during synaptogenesis. The Journal of physiology 539, 53-65 (2002). 156 Munno, D. W., Prince, D. J. & Syed, N. I. Synapse number and synaptic efficacy are regulated by presynaptic cAMP and protein kinase A. The Journal of neuroscience : the official journal of the Society for Neuroscience 23, 4146-4155 (2003). 157 Kehoe, J. Three acetylcholine receptors in Aplysia neurones. The Journal of physiology 225, 115-146 (1972). 158 van Nierop, P. et al. Identification of molluscan nicotinic acetylcholine receptor (nAChR) subunits involved in formation of cation- and anion-selective nAChRs. The Journal of neuroscience : the official journal of the Society for Neuroscience 25, 10617-10626, doi:10.1523/JNEUROSCI.2015-05.2005 (2005).

291

159 van Nierop, P. et al. Identification and functional expression of a family of nicotinic acetylcholine receptor subunits in the central nervous system of the mollusc Lymnaea stagnalis. The Journal of biological chemistry 281, 1680-1691, doi:10.1074/jbc.M508571200 (2006). 160 Price, D. A. & Greenberg, M. J. Purification and characterization of a cardioexcitatory neuropeptide from the central ganglia of a bivalve mollusc. Preparative biochemistry 7, 261-281, doi:10.1080/00327487708061643 (1977). 161 Benjamin, P. R. & Burke, J. F. Alternative mRNA splicing of the FMRFamide gene and its role in neuropeptidergic signalling in a defined neural network. BioEssays : news and reviews in molecular, cellular and developmental biology 16, 335-342, doi:10.1002/bies.950160508 (1994). 162 Skingsley, D. R. et al. A molecularly defined cardiorespiratory interneuron expressing SDPFLRFamide/GDPFLRFamide in the snail Lymnaea: monosynaptic connections and pharmacology. Journal of neurophysiology 69, 915-927 (1993). 163 Santama, N. & Benjamin, P. R. Gene expression and function of FMRFamide-related neuropeptides in the snail Lymnaea. Microscopy research and technique 49, 547-556, doi:10.1002/1097-0029(20000615)49:6<547::AID-JEMT5>3.0.CO;2-Y (2000). 164 Knol, J. C., Weidemann, W., Planta, R. J., Vreugdenhil, E. & van Heerikhuizen, H. Molecular cloning of G protein alpha subunits from the central nervous system of the mollusc Lymnaea stagnalis. FEBS letters 314, 215-219 (1992). 165 Knol, J. C. et al. A novel G protein alpha subunit containing atypical guanine nucleotide- binding domains is differentially expressed in a molluscan nervous system. The Journal of biological chemistry 270, 18804-18808 (1995). 166 Knol, J. C. et al. Cloning of a molluscan G protein alpha subunit of the Gq class which is expressed differentially in identified neurons. European journal of biochemistry / FEBS 230, 193-199 (1995). 167 Fukusumi, S., Fujii, R. & Hinuma, S. Recent advances in mammalian RFamide peptides: the discovery and functional analyses of PrRP, RFRPs and QRFP. Peptides 27, 1073- 1086, doi:10.1016/j.peptides.2005.06.031 (2006). 168 Man-Son-Hing, H., Zoran, M. J., Lukowiak, K. & Haydon, P. G. A neuromodulator of synaptic transmission acts on the secretory apparatus as well as on ion channels. Nature 341, 237-239, doi:10.1038/341237a0 (1989). 169 Haydon, P. G., Man-Son-Hing, H., Doyle, R. T. & Zoran, M. FMRFamide modulation of secretory machinery underlying presynaptic inhibition of synaptic transmission requires a pertussis toxin-sensitive G-protein. The Journal of neuroscience : the official journal of the Society for Neuroscience 11, 3851-3860 (1991). 170 Lovell, P., McMahon, B. & Syed, N. I. Synaptic precedence during synapse formation between reciprocally connected neurons involves transmitter-receptor interactions and AA metabolites. Journal of neurophysiology 88, 1328-1338 (2002). 171 Piomelli, D. et al. Lipoxygenase metabolites of arachidonic acid as second messengers for presynaptic inhibition of Aplysia sensory cells. Nature 328, 38-43, doi:10.1038/328038a0 (1987). 172 Schaechter, J. D. & Benowitz, L. I. Activation of protein kinase C by arachidonic acid selectively enhances the phosphorylation of GAP-43 in nerve terminal membranes. The

292

Journal of neuroscience : the official journal of the Society for Neuroscience 13, 4361- 4371 (1993). 173 Zhang, Z., Bhalla, A., Dean, C., Chapman, E. R. & Jackson, M. B. Synaptotagmin IV: a multifunctional regulator of peptidergic nerve terminals. Nature neuroscience 12, 163- 171, doi:10.1038/nn.2252 (2009). 174 Littleton, J. T., Serano, T. L., Rubin, G. M., Ganetzky, B. & Chapman, E. R. Synaptic function modulated by changes in the ratio of synaptotagmin I and IV. Nature 400, 757- 760, doi:10.1038/23462 (1999). 175 Vician, L. et al. Synaptotagmin IV is an immediate early gene induced by depolarization in PC12 cells and in brain. Proceedings of the National Academy of Sciences of the United States of America 92, 2164-2168 (1995). 176 Chandrasekharappa, S. C. et al. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 276, 404-407 (1997). 177 Guru, S. C. et al. Menin, the product of the MEN1 gene, is a nuclear protein. Proceedings of the National Academy of Sciences of the United States of America 95, 1630-1634 (1998). 178 Huang, J. et al. The same pocket in menin binds both MLL and JUND but has opposite effects on transcription. Nature 482, 542-546, doi:10.1038/nature10806 (2012). 179 Green, D. R. & Kroemer, G. Cytoplasmic functions of the tumour suppressor p53. Nature 458, 1127-1130, doi:10.1038/nature07986 (2009). 180 Cao, Y. et al. Nuclear-cytoplasmic shuttling of menin regulates nuclear translocation of {beta}-catenin. Molecular and cellular biology 29, 5477-5487, doi:10.1128/MCB.00335- 09 (2009). 181 Stewart, C. et al. Characterization of the mouse Men1 gene and its expression during development. Oncogene 17, 2485-2493, doi:10.1038/sj.onc.1202164 (1998). 182 Pfaffl, M. W., Horgan, G. W. & Dempfle, L. Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative expression results in real- time PCR. Nucleic acids research 30, e36 (2002). 183 Schot, L. P. & Boer, H. H. Immunocytochemical demonstration of peptidergic cells in the pond snail Lymnaea stagnalis with an antiserum to the molluscan cardioactive tetrapeptide FMRF-amide. Cell and tissue research 225, 347-354 (1982). 184 Gray, E. G. & Whittaker, V. P. The isolation of nerve endings from brain: an electron- microscopic study of cell fragments derived by homogenization and centrifugation. Journal of anatomy 96, 79-88 (1962). 185 Abrams, H. D., Rohrschneider, L. R. & Eisenman, R. N. Nuclear location of the putative transforming protein of avian myelocytomatosis virus. Cell 29, 427-439 (1982). 186 DuVerle, D. A., Ono, Y., Sorimachi, H. & Mamitsuka, H. Calpain cleavage prediction using multiple kernel learning. PloS one 6, e19035, doi:10.1371/journal.pone.0019035 (2011). 187 Woodin, M. A., Munno, D. W. & Syed, N. I. Trophic factor-induced excitatory synaptogenesis involves postsynaptic modulation of nicotinic acetylcholine receptors. The Journal of neuroscience : the official journal of the Society for Neuroscience 22, 505-514 (2002).

293

188 Luk, C. C. et al. Trophic factor-induced activity 'signature' regulates the functional expression of postsynaptic excitatory acetylcholine receptors required for synaptogenesis. Scientific reports 5, 9523, doi:10.1038/srep09523 (2015). 189 Wu, H. Y. & Lynch, D. R. Calpain and synaptic function. Molecular neurobiology 33, 215-236, doi:10.1385/MN:33:3:215 (2006). 190 Pariat, M. et al. Proteolysis by calpains: a possible contribution to degradation of p53. Molecular and cellular biology 17, 2806-2815 (1997). 191 Kimura, Y. et al. The involvement of calpain-dependent proteolysis of the tumor suppressor NF2 (merlin) in schwannomas and meningiomas. Nature medicine 4, 915-922 (1998). 192 Briz, V. et al. Calpain-2-mediated PTEN degradation contributes to BDNF-induced stimulation of dendritic protein synthesis. The Journal of neuroscience : the official journal of the Society for Neuroscience 33, 4317-4328, doi:10.1523/JNEUROSCI.4907- 12.2013 (2013). 193 Lu, X., Rong, Y. & Baudry, M. Calpain-mediated degradation of PSD-95 in developing and adult rat brain. Neuroscience letters 286, 149-153 (2000). 194 Cohen, S. & Greenberg, M. E. Communication between the synapse and the nucleus in neuronal development, plasticity, and disease. Annual review of cell and developmental biology 24, 183-209, doi:10.1146/annurev.cellbio.24.110707.175235 (2008). 195 Bourtchuladze, R. et al. Deficient long-term memory in mice with a targeted mutation of the cAMP-responsive element-binding protein. Cell 79, 59-68 (1994). 196 Sadamoto, H. et al. CREB in the pond snail Lymnaea stagnalis: cloning, gene expression, and function in identifiable neurons of the central nervous system. Journal of neurobiology 58, 455-466, doi:10.1002/neu.10296 (2004). 197 Yin, J. C. et al. Induction of a dominant negative CREB transgene specifically blocks long-term memory in Drosophila. Cell 79, 49-58 (1994). 198 Deisseroth, K., Bito, H. & Tsien, R. W. Signaling from synapse to nucleus: postsynaptic CREB phosphorylation during multiple forms of hippocampal synaptic plasticity. Neuron 16, 89-101 (1996). 199 Fromaget, M. et al. Functional characterization of a promoter region in the human MEN1 tumor suppressor gene. Journal of molecular biology 333, 87-102 (2003). 200 Murphy, T. H., Worley, P. F. & Baraban, J. M. L-type voltage-sensitive calcium channels mediate synaptic activation of immediate early genes. Neuron 7, 625-635 (1991). 201 Wheeler, D. G. et al. Ca(V)1 and Ca(V)2 channels engage distinct modes of Ca(2+) signaling to control CREB-dependent gene expression. Cell 149, 1112-1124, doi:10.1016/j.cell.2012.03.041 (2012). 202 La, P. et al. Direct binding of DNA by tumor suppressor menin. The Journal of biological chemistry 279, 49045-49054, doi:10.1074/jbc.M409358200 (2004). 203 Colombo, S. F., Mazzo, F., Pistillo, F. & Gotti, C. Biogenesis, trafficking and upregulation of nicotinic ACh receptors. Biochemical pharmacology 86, 1063-1073, doi:10.1016/j.bcp.2013.06.023 (2013). 204 MacConaill, L. E., Hughes, C. M., Rozenblatt-Rosen, O., Nannepaga, S. & Meyerson, M. Phosphorylation of the menin tumor suppressor protein on serine 543 and serine 583. Molecular cancer research : MCR 4, 793-801, doi:10.1158/1541-7786.MCR-06-0123 (2006).

294

205 Bose, C. M. et al. Agrin controls synaptic differentiation in hippocampal neurons. The Journal of neuroscience : the official journal of the Society for Neuroscience 20, 9086- 9095 (2000). 206 Garcia-Osta, A. et al. MuSK expressed in the brain mediates cholinergic responses, synaptic plasticity, and memory formation. The Journal of neuroscience : the official journal of the Society for Neuroscience 26, 7919-7932, doi:10.1523/JNEUROSCI.1674- 06.2006 (2006). 207 Gray, R., Rajan, A. S., Radcliffe, K. A., Yakehiro, M. & Dani, J. A. Hippocampal synaptic transmission enhanced by low concentrations of nicotine. Nature 383, 713-716, doi:10.1038/383713a0 (1996). 208 Genzen, J. R. & McGehee, D. S. Short- and long-term enhancement of excitatory transmission in the spinal cord dorsal horn by nicotinic acetylcholine receptors. Proceedings of the National Academy of Sciences of the United States of America 100, 6807-6812, doi:10.1073/pnas.1131709100 (2003). 209 Miletic, G. & Miletic, V. Increases in the concentration of brain derived neurotrophic factor in the lumbar spinal dorsal horn are associated with pain behavior following chronic constriction injury in rats. Neuroscience letters 319, 137-140 (2002). 210 Phillips, H. S. et al. BDNF mRNA is decreased in the hippocampus of individuals with Alzheimer's disease. Neuron 7, 695-702 (1991). 211 Lamb, P. W., Melton, M. A. & Yakel, J. L. Inhibition of neuronal nicotinic acetylcholine receptor channels expressed in Xenopus oocytes by beta-amyloid1-42 peptide. Journal of molecular neuroscience : MN 27, 13-21, doi:10.1385/JMN:27:1:013 (2005). 212 Spencer, G. E., Lukowiak, K. & Syed, N. I. Transmitter-receptor interactions between growth cones of identified Lymnaea neurons determine target cell selection in vitro. The Journal of neuroscience : the official journal of the Society for Neuroscience 20, 8077- 8086 (2000). 213 Getz, A. M. et al. Two proteolytic fragments of menin coordinate the nuclear transcription and postsynaptic clustering of neurotransmitter receptors during synaptogenesis between Lymnaea neurons. Scientific reports 6, 31779, doi:10.1038/srep31779 (2016). 214 Yokoyama, A. et al. The menin tumor suppressor protein is an essential oncogenic cofactor for MLL-associated leukemogenesis. Cell 123, 207-218, doi:10.1016/j.cell.2005.09.025 (2005). 215 Busygina, V. et al. Hypermutability in a Drosophila model for multiple endocrine neoplasia type 1. Human molecular genetics 13, 2399-2408, doi:10.1093/hmg/ddh271 (2004). 216 Wu, T., Zhang, X., Huang, X., Yang, Y. & Hua, X. Regulation of cyclin B2 expression and cell cycle G2/m transition by menin. The Journal of biological chemistry 285, 18291- 18300, doi:10.1074/jbc.M110.106575 (2010). 217 Gu, Z. & Yakel, J. L. Timing-dependent septal cholinergic induction of dynamic hippocampal synaptic plasticity. Neuron 71, 155-165, doi:10.1016/j.neuron.2011.04.026 (2011). 218 Davies, P. & Maloney, A. J. Selective loss of central cholinergic neurons in Alzheimer's disease. Lancet 2, 1403 (1976).

295

219 Guan, Z. Z., Zhang, X., Ravid, R. & Nordberg, A. Decreased protein levels of nicotinic receptor subunits in the hippocampus and temporal cortex of patients with Alzheimer's disease. Journal of neurochemistry 74, 237-243 (2000). 220 Wonnacott, S. Presynaptic nicotinic ACh receptors. Trends in neurosciences 20, 92-98 (1997). 221 Dutar, P., Bassant, M. H., Senut, M. C. & Lamour, Y. The septohippocampal pathway: structure and function of a central cholinergic system. Physiological reviews 75, 393-427 (1995). 222 Caron, J. et al. Conical diffraction illumination opens the way for low phototoxicity super-resolution imaging. Cell adhesion & migration 8, 430-439, doi:10.4161/cam.29358 (2014). 223 Xu, F. et al. Mercury-induced toxicity of rat cortical neurons is mediated through N- Methyl-D-Aspartate receptors. Molecular brain 5, 30, doi:10.1186/1756-6606-5-30 (2012). 224 La, P. et al. Tumor suppressor menin: the essential role of nuclear localization signal domains in coordinating gene expression. Oncogene 25, 3537-3546, doi:10.1038/sj.onc.1209400 (2006). 225 Couturier, S. et al. A neuronal nicotinic acetylcholine receptor subunit (alpha 7) is developmentally regulated and forms a homo-oligomeric channel blocked by alpha-BTX. Neuron 5, 847-856 (1990). 226 Radcliffe, K. A. & Dani, J. A. Nicotinic stimulation produces multiple forms of increased glutamatergic synaptic transmission. The Journal of neuroscience : the official journal of the Society for Neuroscience 18, 7075-7083 (1998). 227 Reist, N. E., Werle, M. J. & McMahan, U. J. Agrin released by motor neurons induces the aggregation of acetylcholine receptors at neuromuscular junctions. Neuron 8, 865-868 (1992). 228 Garzon, D., Yu, G. & Fahnestock, M. A new brain-derived neurotrophic factor transcript and decrease in brain-derived neurotrophic factor transcripts 1, 2 and 3 in Alzheimer's disease parietal cortex. Journal of neurochemistry 82, 1058-1064 (2002). 229 Webster, S. J., Bachstetter, A. D., Nelson, P. T., Schmitt, F. A. & Van Eldik, L. J. Using mice to model Alzheimer's dementia: an overview of the clinical disease and the preclinical behavioral changes in 10 mouse models. Frontiers in genetics 5, 88, doi:10.3389/fgene.2014.00088 (2014). 230 Guru, S. C. et al. Isolation, genomic organization, and expression analysis of Men1, the murine homolog of the MEN1 gene. Mammalian genome : official journal of the International Mammalian Genome Society 10, 592-596 (1999). 231 Wautot, V. et al. Expression analysis of endogenous menin, the product of the multiple endocrine neoplasia type 1 gene, in cell lines and human tissues. International journal of cancer. Journal international du cancer 85, 877-881 (2000). 232 Debelenko, L. V. et al. Menin immunoreactivity in secretory granules of human pancreatic islet cells. Applied immunohistochemistry & molecular morphology : AIMM / official publication of the Society for Applied Immunohistochemistry 22, 748-755, doi:10.1097/PAI.0000000000000046 (2014).

296

233 Yoshikawa, H., Hellstrom-Lindahl, E. & Grill, V. Evidence for functional nicotinic receptors on pancreatic beta cells. Metabolism: clinical and experimental 54, 247-254, doi:10.1016/j.metabol.2004.08.020 (2005). 234 Boulter, J. et al. Alpha 3, alpha 5, and beta 4: three members of the rat neuronal nicotinic acetylcholine receptor-related gene family form a gene cluster. The Journal of biological chemistry 265, 4472-4482 (1990). 235 Wada, E., McKinnon, D., Heinemann, S., Patrick, J. & Swanson, L. W. The distribution of mRNA encoded by a new member of the neuronal nicotinic acetylcholine receptor gene family (alpha 5) in the rat central nervous system. Brain research 526, 45-53 (1990). 236 McGehee, D. S., Heath, M. J., Gelber, S., Devay, P. & Role, L. W. Nicotine enhancement of fast excitatory synaptic transmission in CNS by presynaptic receptors. Science 269, 1692-1696 (1995). 237 Bertrand, D., Bertrand, S. & Ballivet, M. Pharmacological properties of the homomeric alpha 7 receptor. Neuroscience letters 146, 87-90 (1992). 238 Yu, C. R. & Role, L. W. Functional contribution of the alpha7 subunit to multiple subtypes of nicotinic receptors in embryonic chick sympathetic neurones. The Journal of physiology 509 ( Pt 3), 651-665 (1998). 239 Salas, R. et al. The nicotinic acetylcholine receptor subunit alpha 5 mediates short-term effects of nicotine in vivo. Molecular pharmacology 63, 1059-1066 (2003). 240 Paylor, R. et al. Alpha7 nicotinic receptor subunits are not necessary for hippocampal- dependent learning or sensorimotor gating: a behavioral characterization of Acra7- deficient mice. Learning & memory 5, 302-316 (1998). 241 Bertolino, P. et al. Genetic ablation of the tumor suppressor menin causes lethality at mid-gestation with defects in multiple organs. Mechanisms of development 120, 549-560 (2003). 242 Essrich, C., Lorez, M., Benson, J. A., Fritschy, J. M. & Luscher, B. Postsynaptic clustering of major GABAA receptor subtypes requires the gamma 2 subunit and gephyrin. Nature neuroscience 1, 563-571, doi:10.1038/2798 (1998). 243 Yu, W. et al. Gephyrin interacts with the glutamate receptor interacting protein 1 isoforms at GABAergic synapses. Journal of neurochemistry 105, 2300-2314, doi:10.1111/j.1471-4159.2008.05311.x (2008). 244 Conroy, W. G., Liu, Z., Nai, Q., Coggan, J. S. & Berg, D. K. PDZ-containing proteins provide a functional postsynaptic scaffold for nicotinic receptors in neurons. Neuron 38, 759-771 (2003). 245 Parker, M. J., Zhao, S., Bredt, D. S., Sanes, J. R. & Feng, G. PSD93 regulates synaptic stability at neuronal cholinergic synapses. The Journal of neuroscience : the official journal of the Society for Neuroscience 24, 378-388, doi:10.1523/JNEUROSCI.3865- 03.2004 (2004). 246 Moore, K. A. et al. Partial peripheral nerve injury promotes a selective loss of GABAergic inhibition in the superficial dorsal horn of the spinal cord. The Journal of neuroscience : the official journal of the Society for Neuroscience 22, 6724-6731, doi:20026611 (2002). 247 Campbell, N. R., Fernandes, C. C., Halff, A. W. & Berg, D. K. Endogenous signaling through alpha7-containing nicotinic receptors promotes maturation and integration of

297

adult-born neurons in the hippocampus. The Journal of neuroscience : the official journal of the Society for Neuroscience 30, 8734-8744, doi:10.1523/JNEUROSCI.0931-10.2010 (2010). 248 Liu, Z., Neff, R. A. & Berg, D. K. Sequential interplay of nicotinic and GABAergic signaling guides neuronal development. Science 314, 1610-1613, doi:10.1126/science.1134246 (2006). 249 Lozada, A. F. et al. Glutamatergic synapse formation is promoted by alpha7-containing nicotinic acetylcholine receptors. The Journal of neuroscience : the official journal of the Society for Neuroscience 32, 7651-7661, doi:10.1523/JNEUROSCI.6246-11.2012 (2012). 250 da Penha Berzaghi, M. et al. Cholinergic regulation of brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) but not neurotrophin-3 (NT-3) mRNA levels in the developing rat hippocampus. The Journal of neuroscience : the official journal of the Society for Neuroscience 13, 3818-3826 (1993). 251 Hyman, B. T., Van Hoesen, G. W., Damasio, A. R. & Barnes, C. L. Alzheimer's disease: cell-specific pathology isolates the hippocampal formation. Science 225, 1168-1170 (1984). 252 Geula, C., Nagykery, N., Nicholas, A. & Wu, C. K. Cholinergic neuronal and axonal abnormalities are present early in aging and in Alzheimer disease. Journal of neuropathology and experimental neurology 67, 309-318, doi:10.1097/NEN.0b013e31816a1df3 (2008). 253 Chong, S. A. et al. Synaptic dysfunction in hippocampus of transgenic mouse models of Alzheimer's disease: a multi-electrode array study. Neurobiology of disease 44, 284-291, doi:10.1016/j.nbd.2011.07.006 (2011). 254 Kuchibhotla, K. V. et al. Abeta plaques lead to aberrant regulation of calcium homeostasis in vivo resulting in structural and functional disruption of neuronal networks. Neuron 59, 214-225, doi:10.1016/j.neuron.2008.06.008 (2008). 255 Yin, Y. et al. Accumulation of human full-length tau induces degradation of nicotinic acetylcholine receptor alpha4 via activating calpain-2. Scientific reports 6, 27283, doi:10.1038/srep27283 (2016). 256 Burnstock, G. Do some nerve cells release more than one transmitter? Neuroscience 1, 239-248 (1976). 257 Dale, H. Pharmacology and Nerve-endings (Walter Ernest Dixon Memorial Lecture): (Section of Therapeutics and Pharmacology). Proceedings of the Royal Society of Medicine 28, 319-332 (1935). 258 van den Pol, A. N. Neuropeptide transmission in brain circuits. Neuron 76, 98-115, doi:10.1016/j.neuron.2012.09.014 (2012). 259 Kueh, D. & Jellies, J. A. Targeting a neuropeptide to discrete regions of the motor arborizations of a single neuron. The Journal of experimental biology 215, 2108-2116, doi:10.1242/jeb.067603 (2012). 260 Verhage, M. et al. Differential release of amino acids, neuropeptides, and catecholamines from isolated nerve terminals. Neuron 6, 517-524 (1991). 261 Van Minnen, J. et al. De novo protein synthesis in isolated axons of identified neurons. Neuroscience 80, 1-7 (1997).

298

262 Janes, T. A., Xu, F. & Syed, N. I. Graded hypoxia acts through a network of distributed peripheral oxygen chemoreceptors to produce changes in respiratory behaviour and plasticity. The European journal of neuroscience 42, 1858-1871, doi:10.1111/ejn.12940 (2015). 263 Saunders, S. E., Kellett, E., Bright, K., Benjamin, P. R. & Burke, J. F. Cell-specific alternative RNA splicing of an FMRFamide gene transcript in the brain. The Journal of neuroscience : the official journal of the Society for Neuroscience 12, 1033-1039 (1992). 264 Wang, Z., Lange, A. B. & Orchard, I. Coupling of a single receptor to two different G proteins in the signal transduction of FMRFamide-related peptides. Biochemical and biophysical research communications 212, 531-538 (1995). 265 Sollner, T., Bennett, M. K., Whiteheart, S. W., Scheller, R. H. & Rothman, J. E. A protein assembly-disassembly pathway in vitro that may correspond to sequential steps of synaptic vesicle docking, activation, and fusion. Cell 75, 409-418 (1993). 266 Washbourne, P., Schiavo, G. & Montecucco, C. Vesicle-associated membrane protein-2 (synaptobrevin-2) forms a complex with synaptophysin. The Biochemical journal 305 ( Pt 3), 721-724 (1995). 267 Nakhost, A., Houeland, G., Blandford, V. E., Castellucci, V. F. & Sossin, W. S. Identification and characterization of a novel C2B splice variant of synaptotagmin I. Journal of neurochemistry 89, 354-363, doi:10.1111/j.1471-4159.2004.02325.x (2004). 268 Martina, M. et al. Recordings of cultured neurons and synaptic activity using patch- clamp chips. Journal of neural engineering 8, 034002, doi:10.1088/1741- 2560/8/3/034002 (2011). 269 Johnston, P. A., Jahn, R. & Sudhof, T. C. Transmembrane topography and evolutionary conservation of synaptophysin. The Journal of biological chemistry 264, 1268-1273 (1989). 270 Martin, K. C. et al. Evidence for synaptotagmin as an inhibitory clamp on synaptic vesicle release in Aplysia neurons. Proceedings of the National Academy of Sciences of the United States of America 92, 11307-11311 (1995). 271 Roszer, T. et al. A possible stimulatory effect of FMRFamide on neural nitric oxide production in the central nervous system of Helix lucorum L. Brain, behavior and evolution 63, 23-33, doi:10.1159/000073757 (2004). 272 Kwon, S. E. & Chapman, E. R. Glycosylation is dispensable for sorting of synaptotagmin 1 but is critical for targeting of SV2 and synaptophysin to recycling synaptic vesicles. The Journal of biological chemistry 287, 35658-35668, doi:10.1074/jbc.M112.398883 (2012). 273 Zhu, L. Q. et al. GSK-3 beta inhibits presynaptic vesicle exocytosis by phosphorylating P/Q-type calcium channel and interrupting SNARE complex formation. The Journal of neuroscience : the official journal of the Society for Neuroscience 30, 3624-3633, doi:10.1523/JNEUROSCI.5223-09.2010 (2010). 274 ter Maat, A., Geraerts, W. P., Jansen, R. F. & Bos, N. P. Chemically mediated positive feedback generates long-lasting afterdischarge in a molluscan neuroendocrine system. Brain research 438, 77-82 (1988). 275 Schmidle, T. et al. Synaptin/synaptophysin, p65 and SV2: their presence in adrenal chromaffin granules and sympathetic large dense core vesicles. Biochimica et biophysica acta 1060, 251-256 (1991).

299

276 Obendorf, D., Schwarzenbrunner, U., Fischer-Colbrie, R., Laslop, A. & Winkler, H. In adrenal medulla synaptophysin (protein p38) is present in chromaffin granules and in a special vesicle population. Journal of neurochemistry 51, 1573-1580 (1988). 277 Barkus, R. V., Klyachko, O., Horiuchi, D., Dickson, B. J. & Saxton, W. M. Identification of an axonal kinesin-3 motor for fast anterograde vesicle transport that facilitates retrograde transport of neuropeptides. Molecular biology of the cell 19, 274-283, doi:10.1091/mbc.E07-03-0261 (2008). 278 Adams, D. J., Arthur, C. P. & Stowell, M. H. Architecture of the Synaptophysin/Synaptobrevin Complex: Structural Evidence for an Entropic Clustering Function at the Synapse. Scientific reports 5, 13659, doi:10.1038/srep13659 (2015). 279 Becher, A. et al. The synaptophysin-synaptobrevin complex: a hallmark of synaptic vesicle maturation. The Journal of neuroscience : the official journal of the Society for Neuroscience 19, 1922-1931 (1999). 280 Becher, A., Drenckhahn, A., Pahner, I. & Ahnert-Hilger, G. The synaptophysin- synaptobrevin complex is developmentally upregulated in cultivated neurons but is absent in neuroendocrine cells. European journal of cell biology 78, 650-656, doi:10.1016/S0171-9335(99)80050-8 (1999). 281 Heilig, M. The NPY system in stress, anxiety and depression. Neuropeptides 38, 213- 224, doi:10.1016/j.npep.2004.05.002 (2004). 282 Hinuma, S. et al. New neuropeptides containing carboxy-terminal RFamide and their receptor in mammals. Nature cell biology 2, 703-708, doi:10.1038/35036326 (2000). 283 Krishnan, V. & Nestler, E. J. The molecular neurobiology of depression. Nature 455, 894-902, doi:10.1038/nature07455 (2008). 284 Mei, L. & Nave, K. A. Neuregulin-ERBB signaling in the nervous system and neuropsychiatric diseases. Neuron 83, 27-49, doi:10.1016/j.neuron.2014.06.007 (2014). 285 Tarpey, P. S. et al. A systematic, large-scale resequencing screen of X-chromosome coding exons in mental retardation. Nature genetics 41, 535-543, doi:10.1038/ng.367 (2009). 286 Bauman, J. G. & Bentvelzen, P. Flow cytometric detection of ribosomal RNA in suspended cells by fluorescent in situ hybridization. Cytometry 9, 517-524, doi:10.1002/cyto.990090602 (1988). 287 Borkin, D. et al. Pharmacologic inhibition of the Menin-MLL interaction blocks progression of MLL leukemia in vivo. Cancer cell 27, 589-602, doi:10.1016/j.ccell.2015.02.016 (2015). 288 Conroy, W. G., Vernallis, A. B. & Berg, D. K. The alpha 5 gene product assembles with multiple acetylcholine receptor subunits to form distinctive receptor subtypes in brain. Neuron 9, 679-691 (1992). 289 Girod, R., Barazangi, N., McGehee, D. & Role, L. W. Facilitation of glutamatergic neurotransmission by presynaptic nicotinic acetylcholine receptors. Neuropharmacology 39, 2715-2725 (2000). 290 Girod, R. et al. Heteromeric complexes of alpha 5 and/or alpha 7 subunits. Effects of calcium and potential role in nicotine-induced presynaptic facilitation. Annals of the New York Academy of Sciences 868, 578-590 (1999).

300

291 Kassner, P. D., Conroy, W. G. & Berg, D. K. Organizing Effects of Rapsyn on Neuronal Nicotinic Acetylcholine Receptors. Molecular and cellular neurosciences 10, 258-270, doi:10.1006/mcne.1998.0664 (1998). 292 Zhang, B. et al. Beta-catenin regulates acetylcholine receptor clustering in muscle cells through interaction with rapsyn. The Journal of neuroscience : the official journal of the Society for Neuroscience 27, 3968-3973, doi:10.1523/JNEUROSCI.4691-06.2007 (2007). 293 Williams, B. M. et al. The long internal loop of the alpha 3 subunit targets nAChRs to subdomains within individual synapses on neurons in vivo. Nature neuroscience 1, 557- 562, doi:10.1038/2792 (1998). 294 Fischer, H., Orr-Urtreger, A., Role, L. W. & Huck, S. Selective deletion of the alpha5 subunit differentially affects somatic-dendritic versus axonally targeted nicotinic ACh receptors in mouse. The Journal of physiology 563, 119-137, doi:10.1113/jphysiol.2004.075788 (2005). 295 Fainzilber, M. et al. CRNF, a molluscan neurotrophic factor that interacts with the p75 neurotrophin receptor. Science 274, 1540-1543 (1996). 296 Hamakawa, T. et al. Excitatory synaptogenesis between identified Lymnaea neurons requires extrinsic trophic factors and is mediated by receptor tyrosine kinases. The Journal of neuroscience : the official journal of the Society for Neuroscience 19, 9306- 9312 (1999). 297 Kassabov, S. R. et al. A single Aplysia neurotrophin mediates synaptic facilitation via differentially processed isoforms. Cell reports 3, 1213-1227, doi:10.1016/j.celrep.2013.03.008 (2013). 298 Mellstrom, B., Savignac, M., Gomez-Villafuertes, R. & Naranjo, J. R. Ca2+-operated transcriptional networks: molecular mechanisms and in vivo models. Physiological reviews 88, 421-449, doi:10.1152/physrev.00041.2005 (2008). 299 Selkoe, D. J. Alzheimer's disease is a synaptic failure. Science 298, 789-791, doi:10.1126/science.1074069 (2002). 300 Dougherty, J. J., Wu, J. & Nichols, R. A. Beta-amyloid regulation of presynaptic nicotinic receptors in rat hippocampus and neocortex. The Journal of neuroscience : the official journal of the Society for Neuroscience 23, 6740-6747 (2003). 301 Liu, Q., Xie, X., Lukas, R. J., St John, P. A. & Wu, J. A novel nicotinic mechanism underlies beta-amyloid-induced neuronal hyperexcitation. The Journal of neuroscience : the official journal of the Society for Neuroscience 33, 7253-7263, doi:10.1523/JNEUROSCI.3235-12.2013 (2013). 302 Evans, T. A. et al. BRCA1 may modulate neuronal cell cycle re-entry in Alzheimer disease. International journal of medical sciences 4, 140-145 (2007). 303 Samsom, J. N. & Wong, A. H. Schizophrenia and Depression Co-Morbidity: What We have Learned from Animal Models. Frontiers in psychiatry 6, 13, doi:10.3389/fpsyt.2015.00013 (2015).

301