in synaptic plasticity

(1)

in synaptic plasticity

By

Suleman Hussain

Thesis for the degree of Philosophiae Doctor (PhD) Institute of Basic Medical Sciences

Faculty of Medicine University of Oslo

2021

(2)

Series of dissertations submitted to the Faculty of Medicine, University of Oslo

ISBN 978-82-8377-911-0

reproduced or transmitted, in any form or by any means, without permission.

Cover: Hanne Baadsgaard Utigard.

Print production: Reprosentralen, University of Oslo.

(3)

This thesis is dedicated to my beloved parents

Zakir Hussain and Naseem Akhtar

(4)

(5)

ACKNOWLEDGEMENTS

The present work was carried out at the Laboratory for Synaptic Plasticity, Institute of Basic Medical Sciences, Department of Anatomy, Faculty of Medicine, University of Oslo.

First of all, I thank The God Almighty for providing me with the opportunity to conduct such interesting research and by whose grace I completed this work.

I want to express my deep and profound gratitude to my supervisor Professor Svend Davanger for introducing me to the fascinating field of neuroscience, for being an outstanding supervisor, for exciting discussions and for always believing in me. I owe him thanks for his guidance, ideas, feedback, endless hours of help, flexibility, enthusiasm, encouragement, genuine caring and concern. He has been a tremendous support, and working with him has been a great pleasure for me. The academic freedom and responsibility he has given me have been very educational. I could not have wished for a better supervisor. Without him, this work would never have been accomplished. I look forward to our continued and future scientific collaborations.

Special thanks to Karen Marie Gujord, Grazyna Babinska, Bjørg Riber, Jorunn Knutsen, Johannes Helm and Bashir Hakim for excellent technical assistance, valuable advices and always being helpful and enthusiastic. Thanks to Carina Knudsen for assistance on graphical design and Professor Finn-Mogens Smedja Haug for invaluable help with the immunogold analysis.

I want to express my gratefulness to all the group members of the Laboratory for Synaptic Plasticity for providing a warm, stimulating, friendly and joyful atmosphere in the lab. In particular, Håvard Ringsevjen, Sanjay Aryal, Dinia Saraj, Daniel Lawer Egbenya, Edward Sanders, Tina Günther, Malte Bieler and Camilla Haglerød. I wish to mention especially Professor Sven Ivar Walaas, who was also my co-supervisor, and Professor Ole Petter Ottersen. They have made great intellectual impacts on the lab.

(8)

I would like to thank all the co-authors who contributed to this thesis in various ways.

Also, my colleagues, Professors Eric Rinvik (deceased), Per Holck, Anne Spurkland, Trygve Brauns Leergaard, Mahmood Reza Amiry-Moghaddam, Farrukh Abbas Chaudhry, Niels Christian Danbolt and Erik Dissen deserve my gratitude.

I want to express my appreciation for the Medical Student Research Program. I will especially thank Jarle Breivik, Else Marie Siebke, Lise Sofie Haug Nissen-Meyer and Borghild Arntsen for opening the door to the scientific community. I am grateful to the University of Oslo for providing research facilities.

Special thanks go to all my friends outside the scientific world for providing my life with a lot of fun and for being great companions. I would like to extend particular thank to my good friend Ayub Ali for exciting discussions and valuable advices.

I am deeply and forever indebted to my parents Zakir Hussain and Naseem Akhtar, for giving me the opportunities that have made me who I am. I want to thank them for their endless, unparalleled love, support, help, and encouragement throughout my entire life. I would like to express my sincere gratitude to my brother Altaf Hussain, my sisters Ayesha Hussain, Shazia Hussain, Nazia Hussain and Fozia Hussain for their irreplaceable support, motivation and for always being there for me. Also, warm thanks go to the rest of my family members, especially Attia Fardoos, Mukhtar Ahmed, Muhammad Ilyas, Waqas Ahmed, Adeel Arif, my nieces and my nephews for lighting up my day. Finally, I am grateful to my beloved wife, Khushbakht Naseem, for her continuous support, enthusiasm, patience and for always taking care of me and my interests. The support and encouragement from my family are worth more than I can express on paper. Without all of you, this work would never have been fulfilled.

Oslo, 3 June, 2021 Suleman Hussain

(9)

ABBREVIATIONS

AMPA Alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid Arc Activity-regulated cytoskeleton-associated protein

AZ Active zone

BiFC Bimolecular fluorescence complementation

BoTx Botulinum toxin

CA1 Cornu Ammonis 1

CA2 Cornu Ammonis 2

CA3 Cornu Ammonis 3

CaMKII Calcium calmodulin-dependent protein kinase II CLSM Confocal laser scanning microscopy CNS Central nervous system

DAB Diaminobenzidine

DNA Deoxyribonucleic acid

EC Entorhinal cortex

ER Endoplasmic reticulum

ESPS Excitatory postsynaptic potential fEPSP Field excitatory postsynaptic potential

FRAP Fluorescence recovery after photobleaching FRET Fluorescence resonance energy transfer

GABA Gamma-aminobutyric acid

GRIP Glutamate receptor-interacting protein

H.M. Henry Molaison

HPR Horseradish peroxidase

ISPS Inhibitory postsynaptic potential

KA Kainic acid

KO Knockout

LH Learned helpless

LSAB Labelled streptavidin-biotin

LTD Long-term depression

(10)

LTP Long-term potentiation MDD Major depressive disorder

mEPSCs Miniature excitatory postsynaptic currents

NLH Non-learned helpless

nm Nanometre

NMDA N-methyl-D-aspartate

NSF N-ethylmaleimide-sensitive factor PICK1 Protein interacting with C-kinase 1 PoCy Postsynaptic cytoplasm PoL Postsynaptic lateral membrane

PP Perforant pathway

PreCy Presynaptic cytoplasm PreL Presynaptic lateral membrane

PSD Postsynaptic density

PSD-95 Postsynaptic density protein 95 PVDF Polyvinylidene fluoride

RNA Ribonucleic acid

SAP102 Synapse-associated protein 102 SAP97 Synapse-associated protein 97 SDS Sodium dodecyl sulphate shRNA Short hairpin RNA

SM Sec1/Munc18-like proteins

SNAP Soluble NSF attachment protein SNAP-23 Synaptosomal-associated protein of 23 kDa SNAP-25 Synaptosomal-associated protein of 25 kDa SNAP-29 Synaptosomal-associated protein of 29 kDa SNAP-47 Synaptosomal-associated protein of 47 kDa

SNARE Soluble-N-ethylmaleimide-sensitive factor attachment protein receptor

STX Syntaxin

t-SNARE Target membrane SNARE

TARP Transmembrane AMPA receptor regulatory protein

(11)

TeTx Tetanus toxin

TIRF Total internal reflection fluorescence microscopy TLE Temporal lobe epilepsy

UV Ultraviolet

v-SNARE Vesicle membrane SNARE

VAMP1 Vesicle-associated membrane protein 1 VAMP2 Vesicle-associated membrane protein 2

WB Western blot

WT Wild type

(12)

LIST OF FIGURES

Figure 1: Nerve and glial cells in the CNS ... 18

Figure 2: The rodent hippocampus ... 22

Figure 3: Excitatory synapse... 25

Figure 4: Synaptic transmission ... 27

Figure 5: The SNARE cycle during vesicle docking and fusion ... 29

Figure 6: The presynaptic SNARE complex ... 32

Figure 7: LTP and LTD ... 40

Figure 8: Transmission electron microscope ... 49

Figure 9: Postembedding immunogold procedure... 51

Figure 10: Postsynaptic SNARE complexes ... 86

(13)

LIST OF ARTICLES INCLUDED IN THE THESIS

Article I

Postsynaptic VAMP/synaptobrevin facilitates differential vesicle trafficking of GluA1 and GluA2 AMPA receptor subunits

Hussain S, Davanger S

PLoS One. 2015 October 21; 10(10):e0140868.

doi: 10.1371/journal.pone.0140868. eCollection 2015.

Article II

SNARE protein syntaxin-1 colocalizes closely with NMDA receptor subunit NR2B in postsynaptic spines in the hippocampus

Hussain S, Ringsevjen H, Egbenya DL, Skjervold TL, Davanger S Frontiers in Molecular Neuroscience. 2016 February 05; 9:10.

doi: 10.3389/fnmol.2016.00010. eCollection 2016.

Article III

A possible postsynaptic role for SNAP-25 in hippocampal synapses

Hussain S, Ringsevjen H, Schupp M, Hvalby Ø, Sørensen JB, Jensen V, Davanger S Brain Structure and Function. 2019 March; 224(2):521-532.

doi: 10.1007/s00429-018-1782-2. Epub 2018 October 30.

Article IV

The calcium sensor synaptotagmin 1 is expressed and regulated in hippocampal postsynaptic spines

Hussain S, Egbenya DL, Lai YC, Dosa ZJ, Sørensen JB, Anderson AE, Davanger S Hippocampus. 2017 November; 27(11):1168-1177.

doi: 10.1002/hipo.22761. Epub 2017 July 24.

(14)

Article V

Antibodies raised against aldehyde-fixed antigens improve sensitivity for postembedding electron microscopy

Hussain S, Fredriksen I, Ringsevjen H, Kavalali ET, Davanger S Journal of Neuroscience Methods. 2019 April 01; 317:1-10.

doi: 10.1016/j.jneumeth.2019.01.015. Epub 2019 January 28.

Article VI

Presynaptic PICK1 facilitates trafficking of AMPA-receptors between active zone and synaptic vesicle pool

Haglerød C*, Hussain S*, Nakamura Y, Xia J, Haug FMS, Ottersen OP, Henley JM, Davanger S

Neuroscience. 2017 March 06; 344:102-112.

doi: 10.1016/j.neuroscience.2016.12.042. Epub 2017 January 03.

* Shared first author

Article VII

Changes in concentrations of NMDA receptor subunit GluN2B, Arc and syntaxin-1 in dorsal hippocampus Schaffer collateral synapses in a rat learned helplessness model of depression

Bieler M*, Hussain S*, Daaland ESB, Mirrione MM, Henn FA, Davanger S Journal of Comparative Neurology. 2021 August; 529(12):3194-3205.

doi: 10.1002/cne.25155. Epub 2021 May 27.

* Shared first author

(15)

ADDITIONAL PUBLICATIONS DURING THE PH.D.

PERIOD

Protein interacting with C kinase 1 (PICK1) and GluR2 are associated with presynaptic plasma membrane and vesicles in hippocampal excitatory synapses

Haglerød C, Kapic A, Boulland JL, Hussain S, Holen T, Skare O, Laake P, Ottersen OP, Haug FMS, Davanger S

Neuroscience. 2009 January 12; 158(1):242-52.

doi: 10.1016/j.neuroscience.2008.11.029. Epub 2008 November 27.

The discovery of the soluble N-ethylmaleimide-sensitive factor attachment protein receptor complex and the molecular regulation of synaptic vesicle transmitter release: the 2010 Kavli Prize in neuroscience

Hussain S, Davanger S

Neuroscience. 2011 September 08; 190:12-20.

doi: 10.1016/j.neuroscience.2011.05.057. Epub 2011 June 22.

Changes in synaptic AMPA receptor concentration and composition in chronic temporal lobe epilepsy

Egbenya DL, Hussain S, Lai YC, Xia J, Anderson AE, Davanger S Molecular and Cellular Neuroscience. 2018 October; 92:93-103.

doi: 10.1016/j.mcn.2018.07.004. Epub 2018 July 29.

Omega-3-fatty acids regulate plasticity in distinct hippocampal glutamatergic synapses Aryal S, Hussain S, Drevon CA, Nagelhus E, Hvalby Ø, Jensen V, Walaas SI,

Davanger S

European Journal of Neuroscience. 2019 January; 49(1):40-50.

doi: 10.1111/ejn.14224. Epub 2018 November 19.

(16)

Presynaptic increase in IP3 receptor type 1 concentration in the early phase of hippocampal synaptic plasticity

Ringsevjen H, Hansen HMU, Hussain S, Hvalby Ø, Jensen V, Walaas SI, Davanger S Brain Research. 2019 March 01; 1706:125-134.

doi: 10.1016/j.brainres.2018.10.030. Epub 2018 November 05.

(17)

ABSTRACT

Synaptic plasticity is defined as activity-dependent modification of strength or efficacy of synaptic transmission and entails changes in shape, size, and molecular composition of synapses. Synaptic plasticity is believed to be the cellular basis of learning and memory.

One of the main mechanisms underlying synaptic plasticity is the trafficking of glutamate receptors in and out of synapses. The receptors are inserted into the plasma membrane by fusion of receptor-containing vesicles with the membrane, in a distinct process of exocytosis. However, the molecular mechanisms behind the exocytosis of receptor- containing postsynaptic vesicles are poorly understood. In the presynaptic terminal, assembly of SNARE proteins mediates fusion of vesicles with the target membrane. The presynaptic SNARE core complex consists of vesicle-associated protein VAMP2 and the plasma membrane proteins SNAP-25 and syntaxin-1. The vesicular calcium sensor synaptotagmin 1 interacts with the SNARE complex and triggers exocytosis of synaptic vesicles. The presynaptic localization and function of the SNARE proteins and synaptotagmin 1 are well established. I hypothesized that these proteins are candidates for postsynaptic vesicle fusion as well. The aim of the present thesis was to determine the ultrastructural localization of the SNARE proteins and synaptotagmin 1 in postsynaptic spines, and explore their association with glutamate receptor trafficking. In this thesis, I have demonstrated for the first time the expression of the SNARE proteins and synaptotagmin 1 in postsynaptic spines. The proteins were localized at postsynaptic vesicles and plasma membrane. VAMP2 and synaptotagmin 1 were mainly expressed at postsynaptic cytoplasmic vesicles, while SNAP-25 and syntaxin-1 were present in the highest concentrations at the postsynaptic membrane. Having clarified the presence of the SNARE proteins and synaptotagmin 1 in postsynaptic spines, I have next provided evidence in favour of SNARE-dependent exocytosis of different postsynaptic AMPA- and NMDA-type of glutamate receptors. Moreover, the compositions of postsynaptic SNARE complexes may differ from the presynaptic one. SNARE complexes of different SNARE isoform combinations may possibly co-exist in postsynaptic spines to ensure differential trafficking of glutamate receptors. Neuroplastic pathologies contribute to the development of neuropsychiatric diseases. Thus, in this thesis, I have demonstrated that

(18)

the concentrations of syntaxin-1 and synaptotagmin 1 are changed in animal models of depression and epilepsy, respectively. Taken together, these observations shed light on the postsynaptic role of the SNARE proteins and synaptotagmin 1 in synaptic changes in health and disease.

(19)

INTRODUCTION

The human brain has often been considered as the most complex structure in the universe. It consists of a network of about 86 billion nerve cells assisted by an approximately equal number of glial cells (Azevedo et al., 2009). The nerve cells are interconnected in systems that construct our perception of the external world and control the machinery of our actions.

A nerve cell consists of a cell body, called soma, which contains a nucleus and organelles. The surface area of the cell body is increased by branching projections, called dendrites. Additionally, dendritic processes might contain tiny protrusions known as dendritic spines. An axon is a different projection of a neuronal cell. Compared to dendrites, which tend to be multiple, and highly branched, each neuron has a single long axon that branches at its end with expansions termed nerve terminals or boutons. These terminals in turn form junctions, called synapses, with soma or dendrites from other neurons. Nerve cells vary in size, shape and structure depending on their role and location in the brain (Mark F. Bear, 2007).

In addition to neurons, there are four types of glial cells within the central nervous system (CNS): astrocytes, microglia, oligodendrocytes and ependymal cells. Astrocytes maintain homeostasis, contribute to blood brain barrier and provide support and protection for nerve cells. Astrocytes also communicate with each other, and possibly with neurons as well. Microglia are a specialized population of resident macrophages, they remove cellular debris and mediate immune responses in the CNS (Dale Purves, 2004).

Oligodendrocytes produce insulating myelin sheaths around axons to accelerate neural conduction, whereas ependymal cells are involved in creating cerebrospinal fluid in the central canal of the spinal cord and ventricles of the brain (Kinaan Javed, 2020).

(20)

Figure 1: Nerve and glial cells in the CNS

A diagram showing cellular units of the brain. Figure adapted from (www.dreamstime.com).

Neuronal cells communicate via specialized junctions called synapses. Each neuron can form thousands of junctions with other neurons (Schuz and Palm, 1989). Hence, a typical brain can contain over 100 trillion synapses. An extension of the extracellular space termed the synaptic cleft separates the presynaptic terminal from the postsynaptic spine.

Neuronal communication requires an electrical signal, an action potential, which travels efficiently down the axon and causes the influx of calcium ions into the terminal. This initiates fusion of neurotransmitter-filled presynaptic vesicles with the plasma membrane, in a process termed exocytosis. Synaptic vesicle exocytosis requires specific interaction of the membrane proteins called soluble-N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) in the vesicle and the target membrane (Sollner et al., 1993b).

During exocytosis, the SNARE proteins, vesicle-associated membrane protein 2 (VAMP2, also known as synaptobrevin-2) and the plasma membrane proteins

(21)

synaptosomal-associated protein of 25 kDa (SNAP-25) and syntaxin-1 assemble into the core SNARE complex that drives the vesicular membrane fusion. Synaptotagmin 1 acts as a calcium sensor in the presynaptic terminal that triggers calcium-dependent exocytosis of vesicles by interacting with the SNARE proteins (Sudhof and Rizo, 2011).

In excitatory synapses, exocytosis of vesicles leads to release of neurotransmitter glutamate into the synaptic cleft, which then binds to postsynaptic glutamate receptors, thus conveying the signal from one neuron to another.

One of the most fascinating properties of the brain is its capacity to change functionally and structurally, thereby enabling the organism to adapt to changes in the environment.

This amazing, unique and remarkable ability of the brain to reorganize itself is known as brain plasticity. At the cellular and synaptic levels, plasticity involves changes in the efficacy, size, shape and molecular composition of synapses. Besides, synaptic plasticity is widely believed to be the basis of learning and memory (Citri and Malenka, 2008, Langille and Brown, 2018). Long-term potentiation (LTP) and long-term depression (LTD) are major forms of synaptic plasticity. LTP increases strength of synaptic transmission (Bliss and Lomo, 1973). On the other hand, synapses become weaker during LTD (Luscher and Malenka, 2012). One of the main mechanisms underlying synaptic plasticity is insertion and removal of glutamate receptors from the postsynaptic plasma membrane (Collingridge et al., 2004, Chater and Goda, 2014, Lau and Zukin, 2007). But still, the molecular mechanisms regulating the trafficking of glutamate receptors have not been finally clarified in detail. Understanding synaptic plasticity requires insight into the molecular organization of the glutamatergic synapse and the molecular interactions that underlie glutamate receptor trafficking.

Based on what we know of the presynaptic function of the SNARE proteins, these could be important candidates for the fusion of receptor-containing organelles with the postsynaptic plasma membrane as well. Nevertheless, the composition of the postsynaptic SNARE machinery and its interaction with the glutamate receptors is still poorly understood. So far, there is little information concerning postsynaptic vesicles that must be responsible for the delivery of glutamate receptors to the cell membrane of the

(22)

synapses. No one has yet managed to isolate them in a way that makes it possible to determine in detail proteins they are associated with, or how exocytosis is regulated.

Furthermore, localization of SNARE proteins at the ultrastructural level in the postsynaptic spine has not been demonstrated before the present work. In my thesis, I have not studied synaptic plasticity per se. However, I have determined the localization of the SNARE proteins relative to postsynaptic vesicles and the plasma membrane, and investigated the association of the SNARE proteins with glutamate receptor trafficking within the postsynaptic spine. These are important cell biological factors in synaptic plasticity. This knowledge will lead to a better understanding of the molecular mechanisms underlying normal synaptic function and synaptic changes. The following review will introduce a simplified presentation of some of the most important structural and functional aspects that form the scientific basis of this thesis.

(23)

The hippocampus

The hippocampus is part of the temporal lobe and plays an essential role in learning and memory. The first conclusive evidence for the involvement of the hippocampus in forming new memories was the study of patient Henry Molaison (H.M.) (Scoville and Milner, 1957), who is probably the best-known single patient in the history of neuroscience. Patient H.M. had incurable epilepsy. To relieve his symptoms, bilateral temporal lobe resection was carried out by neurosurgeon William Scoville. The epilepsy was controlled after the surgery. The removal of both hippocampi, however, resulted in the loss of H.M.s ability to form most types of memories (Squire, 2009), indicating that the hippocampus performs a vital role in memory formation.

The hippocampus proper, Cornu Ammonis (CA) is one of several related brain regions that together comprise a functional system called the hippocampal formation. The other regions of the hippocampal formation include the dentate gyrus and subiculum, some also include the presubiculum, parasubiculum and entorhinal cortex. The principal neurons in the dentate gyrus and the hippocampus proper are, respectively, the granule and the pyramidal cells. The dentate gyrus consists of the following three layers: the outer molecular layer, middle granular layer and inner polymorphic cell layer. The hippocampus proper forms the area between the dentate gyrus and the subiculum. It is subdivided into three subregions: CA1, CA2 and CA3. These three are comprised of the following cell layers from the dentate gyrus to the cortex: stratum lacunosum- moleculare, stratum radiatum, stratum pyramidale, stratum oriens and alveus. In the CA3 field, but not CA1 or CA2, the stratum lucidum is located between the stratum radiatum and the pyramidal cell layer (Per Andersen, 2007).

The entorhinal cortex is the primary input and output structure of the hippocampal formation. Cells in the entorhinal cortex give rise to axons that form excitatory connections with granule cells of the dentate gyrus. The projection from the entorhinal cortex to the dentate gyrus is called the perforant pathway. The perforant path fibers do also terminate directly in the CA3. Next, the axons from granule cells form the mossy fiber pathway that connects with pyramidal cells of the CA3 field of the hippocampus.

(24)

The pyramidal axons of the CA3 region, in turn, project through Schaffer collaterals in the stratum radiatum to the pyramidal cells in the CA1. The pyramidal cell axons of the CA1 region finally projects to the subiculum and the entorhinal cortex and completes the intrinsic hippocampal circuit (Per Andersen, 2007).

Figure 2: The rodent hippocampus

A) An illustration over the hippocampal circuitry and the three main regions of the rodent hippocampus: CA1, CA3 and the dentate gyrus. B) Diagram of the hippocampal neural network.

Abbreviations: Perforant pathway (PP); Entorhinal cortex (EC). Roman numerals indicate entorhinal cortical layers. Modified figure from (Deng et al., 2010).

(25)

In the present study, I have focused on the hippocampus, with emphasis on the Schaffer collateral pathway and their glutamatergic synapses in the stratum radiatum of the CA1 region. The cells of the pyramidal layer have dendrites branching out in both stratum oriens and stratum radiatum. The stratum oriens contains the basal dendrites, while stratum radiatum comprises Schaffer collaterals connecting to the apical dendrites. The pyramidal cell layer of the CA2 and CA3 regions is loosely arranged. In contrast, the neurons are tightly packed in the CA1 region of the hippocampus. Furthermore, CA1 pyramidal cells tend to be smaller and more homogenous than in the CA3 and proximal CA2 region (Pyapali et al., 1998). The pyramidal neurons throughout the CA1 have a mean total dendritic length of 13,424 μm (Ishizuka et al., 1995). Moreover, a single pyramidal cell receives around 30 000 excitatory and 1700 inhibitory inputs (Megias et al., 2001).

The anatomical organization of the synapse

Individual axons weaving throughout the hippocampus and elsewhere in the brain make connections and communicate with other neuronal cells via specialized junctions known as synapses. These synapses are classified as chemical or electrical based on whether signalling through the synapses occurs by the release of chemical messengers or direct electrical coupling. Electrical synapses will not be discussed further. Chemical synapses are divided into asymmetric (type 1) and symmetric (type 2) synapses (Peters and Palay, 1996). Asymmetric synapses are excitatory and mainly found on dendrites and dendritic spines. Conversely, symmetric synapses are inhibitory and located on the cell soma and initial segment of axons. A typical excitatory synapse consists of a presynaptic bouton or terminal and a postsynaptic spine. The presynaptic bouton varies in size and contains spherical vesicles of approximately 35 nm in diameter filled with signal molecules (Qu et al., 2009). Presynaptic boutons can also contain other vesicles of diverse shape, size and content. Inhibitory presynaptic boutons contain slightly smaller vesicles, often elongated in shape. An active zone is a specialized region on the presynaptic plasma membrane for exocytosis. Synaptic vesicles close to the active zone are docked and primed for release.

Several evolutionarily conserved proteins form a single large protein complex that makes

(26)

the core of the active zone (Sudhof, 2012). In electron micrographs, the active zone can be recognized by increased electron density of the presynaptic membrane.

The presynaptic terminal is separated from the postsynaptic spine by a 20 nm wide synaptic cleft, which contains standard extracellular matrix proteins and transsynaptic adhesion proteins that span the cleft and connect the presynaptic terminal to the postsynaptic element. Postsynaptic spines are small membranous protrusions from neuronal dendrites. Spines vary greatly in their shape and size in the different brain regions and even along short segments of a single dendrite. A postsynaptic density (PSD) is a prominent electron-dense region in postsynaptic spines located directly opposite the presynaptic active zone. The PSD consists of a multiprotein assembly of receptors, scaffolding proteins and signalling complexes involved in transduction in the postsynaptic cell. It extends 35-50 nm into the cytoplasm beneath the plasma membrane of excitatory synapses (Harris and Weinberg, 2012). Inhibitory synapses lack a prominent PSD. Some spines contain a spine apparatus, which is a derivate from the dendritic smooth endoplasmic reticulum (ER). The spine apparatus is involved in the synthesis of proteins, posttranslational modification of proteins and regulation of calcium (Spacek and Harris, 1997).

(27)

Figure 3: Excitatory synapse

A) A schematic diagram of subregions of an excitatory synapse. B) Electron micrograph of an excitatory synapse. Abbreviations: Active zone (AZ); Presynaptic lateral membrane (PreL);

Presynaptic cytoplasm (PreCy); Postsynaptic density (PSD); Postsynaptic lateral membrane (PoL); Postsynaptic cytoplasm (PoCy). Modified electron micrograph adapted from (www.medcell.med.yale.edu).

Synaptic transmission

Neuronal cells communicate via a combination of electrical impulses and chemical signals. When an electrical impulse, called an action potential, arrives at the presynaptic terminal, it activates presynaptic voltage-gated calcium channels. The rapid rise in intracellular calcium ions triggers synaptic vesicle fusion with the plasma membrane and release of chemical messenger molecules referred as neurotransmitters into the synaptic cleft (Mark F. Bear, 2007). Neurotransmitters are small molecules, such as amino acids, monoamines or peptides. The amino acid glutamate is the primary excitatory

(28)

neurotransmitter at asymmetric synapses, instead the amino acids glycine and gamma- aminobutyric acid (GABA) are the main inhibitory neurotransmitters at symmetric synapses in the CNS (Dale Purves, 2004). The transmitter diffuses to the postsynaptic membrane, binds to postsynaptic receptors and thereby transfers the information to the postsynaptic cell. The effect on the postsynaptic cell depends on the type of receptors present in the postsynaptic plasma membrane. There are two main types of neurotransmitter receptors, metabotropic and ionotropic receptors. Metabotropic receptors are G-protein-coupled receptors that act through a second messenger signalling pathway. Ionotropic receptors are ligand-gated ion channels, which allow ions to pass through the membrane and alter the postsynaptic membrane potential (Hollmann and Heinemann, 1994). The postsynaptic potential can either be an excitatory postsynaptic potential (ESPS), which increases the probability of generating an action potential, or an inhibitory postsynaptic potential (IPSP) that decreases the probability of firing an action potential in the postsynaptic neuron. If the sum of all EPSPs and IPSPs results in a depolarization of the membrane potential sufficient to reach the threshold level, the cell will produce an action potential (Mark F. Bear, 2007).

The action of neurotransmitters is terminated by their removal from the synaptic cleft by diffusion, enzymatic degradation, or by reuptake in neurons or astrocytes. Glutamate is transported into the surrounding astrocytes by excitatory amino acid transporters. Inside the astrocyte, glutamate is converted to glutamine by glutamine synthetase. Glutamine is released from astrocytes to the extracellular space and then taken into the presynaptic nerve terminal, where it is converted to glutamate by phosphate-activated glutaminase.

Finally, glutamate is packed into vesicles by vesicular glutamate transporters (Bak et al., 2006). GABA is also converted to glutamine in astrocytes and released to the extracellular space. Glutamine is then taken up by inhibitory neurons. Inside the neuron, the glutamine is first converted to glutamate and then further metabolized into GABA by glutamate decarboxylase (Bak et al., 2006).

(29)

Figure 4: Synaptic transmission

An illustration of synaptic transmission at a chemical synapse. 1: An action potential arrives at the presynaptic terminal. 2: Presynaptic voltage-gated calcium channels are activated. 3:

Increased calcium concentration in the terminal initiates exocytosis of transmitter-containing vesicles. 4: Transmitter binds to postsynaptic ligand-gated channels. 5: Channels open. 6: Flow of sodium ions depolarizes the postsynaptic neuron. 7: If depolarization reaches the threshold, the neuron will fire an action potential. Figure retrieved from (www.teaching.ncl.ac.uk).

Vesicle exocytosis

Every cell is endowed with the capacity to export internal substances out of the cell and insert molecules into the plasma membrane. In fact, this is one of several defining characteristics of a living cell or organism. One specialized way of achieving this is the fusion of intracellular vesicles with the plasma membrane, in a distinct process of exocytosis. Trafficking of vesicles and exocytosis at the synapse is dependent on specific target membrane recognition, attachment and fusion. Many proteins are involved in this

(30)

process. Among the components of this machinery are SM proteins, SNARE proteins, RIM proteins, synaptotagmins, complexins, Munc13, synuclein chaperones, small GTPases from the Rab family and chaperone complexes composed of CSPα, Hsc70, and SGT (Sudhof and Rizo, 2011). Here the focus will be on the SNARE proteins and its interacting partner synaptotagmin 1.

Rab proteins are a family of small GTPases. They associate with vesicles and target vesicles to their site of fusion by binding to the Rab effectors on the target membrane (Lin and Scheller, 2000). The next step is vesicle docking and fusion, which requires specific interaction of SNARE proteins in the vesicle and target membranes (Sollner et al., 1993b). SNARE proteins constitute a large protein superfamily comprising more than 60 members in both mammalian and yeast cells. Most of these proteins are exposed in the cytoplasm. Only a few SNAREs contain lipid anchors (Sollner, 2003). SNAREs located on the vesicles are classified as vesicle membrane SNAREs (v-SNARE), and SNAREs located on the target membrane is classified as target membrane SNAREs (t-SNARE).

The v- and t-SNARE classification is functional for fusion reactions such as neurotransmitter release that involves fusion between a vesicle and the plasma membrane, but the v- and t-SNARE classification might be confusing for other reactions, such as the homotypic fusion of two symmetric yeast vacuoles. All SNARE proteins share characteristic SNARE motifs of about 60 amino acids including either glutamine (Q) or arginine (R) in the middle. Accordingly, a second nomenclature that categorizes SNAREs into Q- and R-SNAREs has been developed (Fasshauer et al., 1998). Usually, the R-SNARE is contributed by the vesicle and acts as a v-SNARE, and three Q- SNAREs are contributed by the target organelle and act as t-SNAREs.

Membrane fusion is initiated and carried out after the v-SNARE of a membrane has paired together with its cognate t-SNARE on another membrane, thereby forming SNAREpins or trans-SNARE complexes that bridge the gap between the vesicle and the plasma membrane. The assembly of SNARE proteins require Sec1/Munc18-like proteins (SM), which also promote membrane fusion process together with SNARE complex (Sudhof and Rothman, 2009). A structural SNARE complex consists of a four-helix

(31)

bundle where the v-SNARE contributes one-helix and t-SNAREs contributes three helices (Sutton et al., 1998). Protein folding during the formation of a SNAREpin brings the two membranes in close apposition, so that a membrane fusion event may occur (Weber et al., 1998). SNARE assembly releases enough energy to overcome the repulsive forces that usually keep membranes from converging (Fasshauer et al., 1997). After membrane fusion has occurred, the SNARE complex will accumulate in the cis conformation, where all SNARE complexes are in the target membrane. N- ethylmaleimide-sensitive factor (NSF) and alpha soluble NSF attachment protein (SNAP) are cytosolic proteins, which interact with intracellular SNARE complexes. They utilize the energy of ATP hydrolysis to disassemble SNARE complexes and allows the recycling of SNAREs that have finished one round of membrane fusion (Jahn and Sudhof, 1999).

Figure 5: The SNARE cycle during vesicle docking and fusion

Free SNARE proteins at the plasma membrane assemble into acceptor complexes. This assembly requires SM proteins. The acceptor complex interacts with the SNARE protein on the vesicle and forms a four-helical-trans complex, which proceeds from a loose to a tight state through a zippering process. This is followed by opening of a fusion pore. These transitions are controlled by late regulatory proteins that include synaptotagmin and complexin. During fusion, the trans- configuration is converted to cis-configuration. Cis-complex is disassembled by NSF and alpha SNAP by utilizing the energy of ATP hydrolysis. Figure adapted from (Jahn and Scheller, 2006).

(32)

The presynaptic SNARE core complex and SNARE family members

As mentioned above, presynaptic vesicle exocytosis and neurotransmitter release are dependent on SNARE-mediated fusion of synaptic vesicles with the plasma membrane.

The SNARE complex in the presynaptic terminal consists of a v-SNARE, VAMP2, which resides on the synaptic vesicle membrane, and two t-SNARES, syntaxin-1 and SNAP-25 on the plasma membrane (Hussain and Davanger, 2011, Sudhof and Rothman, 2009). The proteins form an extremely stable four-helix complex. VAMP2 and syntaxin- 1 contribute one helix each, by contrast SNAP-25 contributes two helices to the SNARE complex. Synaptic SNARE proteins are the substrate for the proteolytic action of botulinum toxin (BoTx) and tetanus toxin (TeTx), which potently block exocytosis in nerve terminals (Blasi et al., 1993a, Blasi et al., 1993b, Schiavo et al., 1992, Schiavo et al., 1993).

VAMPs are vesicle-associated proteins. There are seven known proteins in the VAMP protein family. Rat brain contains two VAMP isoforms, termed VAMP1 and VAMP2.

The proteins are expressed in a distinct, but slightly overlapping pattern. Although VAMP2 is much more ubiquitously expressed in the brain, VAMP1 is expressed at a higher level in the spinal cord. Both proteins are mainly localized in nerve terminals (Raptis et al., 2005, Elferink et al., 1989). BoTx B and TeTx cleaves VAMP2. VAMP1, however, is insensitive to these toxins (Schiavo et al., 2000).

The SNAP-25 protein family is comprised of SNAP-25, SNAP-23, SNAP-29 and SNAP- 47 (Kadkova et al., 2019). SNAP-25 does not contain a transmembrane domain. The protein is anchored to the cytosolic face of the membrane via palmitoyl side chains.

SNAP-25 exists in two isoforms that differ by only nine amino acids and are referred to as SNAP-25A and SNAP-25B (Bark and Wilson, 1994). SNAP-25A is expressed at the embryonic stage, while SNAP-25B is the major isoform during the postnatal life (Bark et al., 1995). SNAP-25 is widely distributed in the adult rodent brain, localized to axonal processes and nerve terminals (Oyler et al., 1992). BoTx A and E selectively cleaves SNAP-25 (Schiavo et al., 1993). Similar to SNAP-25, SNAP-23 is also anchored to the membrane by palmitoyl side chains. SNAP-23 shows a distinct pattern of expression in

(33)

the brain. The protein is most strongly expressed in the cerebellum, the hippocampus and the cortex, primarily concentrated in cell bodies (Chen et al., 1999). SNAP-27 and SNAP-29 lack palmitoylation sites for membrane anchoring. SNAP-47 is ubiquitously expressed in the brain, located in both axonal and dendritic compartments (Holt et al., 2006, Munster-Wandowski et al., 2017), whereas SNAP-29 is mainly localized at intracellular membranes and only occasionally at the plasma membrane (Steegmaier et al., 1998). The protein is found in low numbers on synaptic vesicles (Takamori et al., 2006).

Syntaxins are transmembrane proteins that consist of a family of 16 members (Bock and Scheller, 1997). From which only four isoforms, syntaxin-1 to syntaxin-4 are localized to the plasma membrane where they mediate membrane fusion (Teng et al., 2001).

Syntaxin-1 exists in two isoforms, syntaxin-1A and syntaxin-1B. They share 84% amino acid identity (Bennett et al., 1992). Both isoforms show an overlapping pattern in most areas of the brain. Despite that, the isoforms are expressed differently across the CNS, like, e.g., the thalamus, the pituitary gland and the spinal cord, where syntaxin-1A is more strongly expressed. Both isoforms are mainly present in axons and boutons (Ruiz- Montasell et al., 1996). BoTx C cleaves syntaxin-1, in addition to SNAP-25 (Blasi et al., 1993b, Williamson et al., 1996). Like syntaxin-1, Syntaxin-4 is also widely expressed in the brain. In contrast, syntaxin-3 expression is highest in the cerebellum and the striatum (Chen et al., 1999). Syntaxin-3 is abundant in axons and terminals (Soo Hoo et al., 2016), while syntaxin-4 is localized in the cell body and dendrites (Kennedy et al., 2010). To my knowledge, there are no studies to date, which have addressed localization of syntaxin-2 in the brain.

Synaptotagmin 1

The calcium signals controlling the exocytosis of vesicles in the presynaptic terminal must at some point be coupled to the exocytotic machinery. In the nerve terminal, synaptotagmin 1, a member of the evolutionarily conserved family of 17 membrane‐

trafficking proteins, is localized to synaptic vesicles and act as a calcium sensor.

(34)

Synaptotagmin 1 contains two C2 domains that bind calcium. The protein mediates calcium-dependent binding of synaptotagmin 1 to the plasma membrane and the assembled SNARE complex (Sollner et al., 1993a). Moreover, synaptotagmin 1 competes with complexin for binding to the SNARE proteins. Complexins are small soluble proteins that also bind to the assembled SNARE complex and block the fusion of vesicles with the cell membrane (Sudhof, 2013). The binding of calcium ions to synaptotagmin 1 during synaptic transmission causes the release of complexin clamp from the SNARE complex. Consequently, synaptotagmin 1 interacts with the SNARE complex and membrane phospholipids, which in turn triggers exocytosis of vesicles (Sudhof, 2013, Sudhof, 2004, Sudhof and Rothman, 2009).

Figure 6: The presynaptic SNARE complex

A) Structure of the presynaptic SNARE complex. B) A model for calcium-dependent vesicle fusion. 1: Vesicles docks. 2: The SNARE proteins interact to bring the membranes together.

3: Influx of calcium. 4: Calcium bound synaptotagmin catalyzes membrane fusion. Figure retrieved from (Dale Purves, 2004).

(35)

Glutamate receptors

Glutamate released from the presynaptic terminal in excitatory synapses diffuses across the synaptic cleft and binds to postsynaptic glutamate receptors, thus transmitting a signal from one neuron to another. Glutamate receptors are categorized as either metabotropic or ionotropic. Metabotropic glutamate receptors are a family of G-protein coupled receptors that participate in the modulation of synaptic transmission and neuronal excitability throughout the CNS. These receptors can be classified into eight subtypes and three subgroups based on their signal transduction pathways and pharmacological profiles (Crupi et al., 2019). Ionotropic glutamate receptors are non-selective cation channels, which mediate fast excitatory neurotransmission in the brain. These receptors are classified according to their pharmacology, i.e., their binding preference for one of the following agonists: N-methyl-D-aspartate (NMDA), alpha-amino-3-hydroxy-5-methyl-4- isoxazolepropionic acid (AMPA) or kainic acid (KA) (Hollmann and Heinemann, 1994).

Kainate receptors are widely expressed pre- and postsynaptically. The receptors modulate both glutamatergic and GABAergic synaptic transmission in a number of brain regions.

The receptors have also been reported to mediate neurotransmission through metabotropic signalling cascades, in addition to their conventional function as ionotropic receptors (Contractor et al., 2011). Kainate receptors are tetramers made up of combinations of five subunits, GluK1-5. Subunits GluK1-3 can form homomeric and heteromeric receptors. GluK4-5, however, require heteromeric assembly with one of the GluK1-3 subunits to form functional channels (Contractor et al., 2011).

AMPA receptors are expressed throughout the brain and mediate the majority of fast excitatory synaptic transmission. The receptors are enriched at excitatory glutamatergic synapses, where they are located at the postsynaptic membrane and serve as key determinants of synaptic strength (Chater and Goda, 2014). Nonetheless, we have also localized such receptors in the presynaptic terminal (Haglerod et al., 2009). AMPA receptors are essential for excitatory synapse formation, stabilization, synaptic plasticity and neural circuit formation (Henley and Wilkinson, 2016). AMPA receptors are tetramers comprising various combinations of the four subunits GluA1-4 that are

(36)

highly homologous. The receptors are expressed as both homomers and heteromers comprising at least two subunits (Wenthold et al., 1996). The GluA4 subunit is mainly expressed during early development and only present at low levels in the adult brain.

GluA2-containing receptors, however, predominate throughout the brain (Isaac et al., 2007), hence the majority of receptors are heterotetramers of GluA1/GluA2 or GluA2/GluA3 subunits (Shepherd and Huganir, 2007).

NMDA receptors are widely distributed in the brain. The majority of NMDA receptors are located postsynaptically on dendrites and dendritic spines. The receptors are crucial for synaptic development, synaptic plasticity and excitotoxicity (Lau and Zukin, 2007). NMDA receptors are also heteromeric molecules, being formed of GluN1, GluN2, or GluN3 subunits. There is a single GluN1 subunit, with eight splice variants, four GluN2 subunits (GluN2A–D) and two GluN3 subunits (GluN3A–B). A functional NMDA receptor is a heterotetramer composed mainly of two GluN1 subunits and two GluN2 subunits. The GluN3 subunit can replace one or both of the GluN2 subunit in the tetramer (Vyklicky et al., 2014).

Glutamate receptor trafficking

Trafficking of glutamate receptors in and out of synaptic sites is considered to be involved in activity-induced changes in synaptic strength. AMPA receptors assemble in the ER first as dimers, which then come together to form dimers of dimers to make a tetramer (Tichelaar et al., 2004). GluA2-containing AMPA receptors have an arginine residue in the pore region, which functions both to act as an ER retention motif and to render GluA2-containing AMPA receptors impermeable to calcium (Greger et al., 2003).

GluA1 lacks this residue, as a result, GluA1 homomers are calcium permeable and is rapidly exported from the ER. Another factor that determines the trafficking of AMPA receptors is the length of intracellular C-terminal domains (tail) of receptor subunits.

GluA1 and GluA4 are long-tailed, contrarily, GluA2 and GluA3 are short-tailed. Subunits with a long domain are rapidly mobilized from the ER to the surface. Short-tailed subunits are trafficked more slowly from the ER. Receptors containing both short- and

(37)

long-tailed subunit combinations exhibit the trafficking properties of long-tailed subunits (Henley and Wilkinson, 2013). The NMDA receptor subunits GluN1 are produced in the ER in large excess relative to GluN2 subunits, ensuring that sufficient amounts of GluN1 subunits are available for newly synthesized GluN2 and GluN3 subunits (Horak et al., 2014). Some studies suggest that GluN1-GluN1 and GluN2-GluN2 homodimers are initially formed before they are further assembled into functional heterotetramers, while others propose that GluN1-GluN2 heterodimers are required for formation of heterotetrameric receptors (Horak et al., 2014).

The exit of AMPA and NMDA receptors from the ER is postulated to need PDZ protein interaction. PDZ domain-containing proteins are a group of proteins that share a domain of 80-90 amino acids. These proteins anchor transmembrane proteins and play a central role in synaptic membrane protein localization (Groc and Choquet, 2006). Newly synthesized glutamate receptors have to travel long distances along dendrites to reach the most distal synapses. The transport of membrane organelles on microtubule tracks is an active process that is depended on the motor proteins of the kinesin and dynein super- families (Groc and Choquet, 2006). Synaptic targeting of the glutamate receptors is a complex process regulated by many proteins, in particular protein kinases and PDZ- domain proteins. Key proteins involved in the trafficking of NMDA and AMPA receptors are SAP102 and PSD-95, which bind to NMDA receptors (Ronald S. Petralia, 2008).

GRIP and NSF interact with GluA2, and finally, SAP97 and protein 4.1N bind to GluA1 (Anggono and Huganir, 2012). These interacting complexes play an essential role in the trafficking and stabilization of glutamate receptors at the synapse. Several glutamate receptor interacting proteins recruit protein kinases and thereby modulate phosphorylation of subunits. This is an important mechanism for activity-dependent trafficking of the receptors (Barry and Ziff, 2002). Directional movement of AMPA receptors towards synapses may be controlled by an exocyst complex component, Sec8, through PDZ-dependent interactions (Gerges et al., 2006). Additionally, members of transmembrane AMPA receptor regulatory protein (TARP) family, e.g., stargazin helps to anchor AMPA receptors in the PSD (Ziff, 2007).

(38)

One of the last steps in the transport of glutamate receptors to the synapse is the insertion of the receptors into the plasma membrane. Receptors can either be directly exocytosed at the synapse or first exocytosed into the extrasynaptic membrane, followed by lateral diffusion to the synaptic site where they are stabilized by synaptic scaffold proteins (Chater and Goda, 2014, Ronald S. Petralia, 2008, Groc and Choquet, 2006, Kennedy and Ehlers, 2011, Henley et al., 2011).

Synaptic trafficking of glutamate receptors is a highly dynamic and tightly regulated process, which involves exocytosis, endocytosis and membrane lateral diffusion of the receptors. The mechanism underlying exocytosis of receptors is not fully understood.

Insertion of AMPA receptor into the plasma membrane is blocked by neurotoxins, suggesting that AMPA receptors are inserted via SNARE-dependent mechanism (Lu et al., 2001, Lledo et al., 1998, Kennedy and Ehlers, 2011). Furthermore, Gerges and colleagues have reported that an exocyst complex component, Exo70, mediates AMPA receptor insertion directly within the PSD (Gerges et al., 2006). In another study, they showed that Rab8 is required for AMPA receptor insertion to the spine plasma membrane during LTP and constitutive cycling (Gerges et al., 2004). Similarly, the role of complexin (Ahmad et al., 2012) and synaptotagmins (Wu et al., 2017) has also been proposed in exocytosis of AMPA receptors. Interactions between SAP102 and exocyst protein, Sec 8, are involved in the delivery of NMDA receptors to the cell surface (Sans et al., 2003). The role of SNARE proteins is also indicated in the insertion of NMDA receptors into the plasma membrane (Suh et al., 2010, Lau et al., 2010, Gu et al., 2016).

Receptors inserted by exocytosis into the synaptic plasma membrane could either come from a pool of newly synthesized receptors or from a pool of recycling receptors (Park, 2018). Glutamate receptors are internalized through a clathrin-mediated mechanism. AP- 2 adaptor complex regulates the endocytosis of NMDA receptors (Lavezzari et al., 2004).

Protein interacting with C-kinase 1 (PICK1) is a synaptic PDZ domain protein, which interacts with the GluA2 subunit and takes part in the NMDA-induced internalization of AMPA receptors (Xia et al., 1999, Hanley and Henley, 2005). Likewise, a member of the

(39)

immediate-early gene family, the activity-regulated cytoskeleton-associated protein (Arc) mediates endocytosis of AMPA receptors (Chowdhury et al., 2006).

Synaptic plasticity

Increased or decreased communication between neurons can induce changes in size, shape, efficiency and molecular composition of their synapses. This ability of synapses to strengthen or weaken over time is termed synaptic plasticity (Citri and Malenka, 2008).

The hypothesis of activity-dependent synaptic plasticity was first introduced by Santiago Ramon y Cajal in 1894. He postulated that the brain can store information by strengthening of existing neuronal connections (Cajal, 1894). This theory was later revised by the Canadian psychologist Donald Hebb, who proposed that neurons strengthen their communication if the presynaptic cell persistently stimulates the postsynaptic cell (Hebb, 1949). This idea is often summarized as “neurons that fire together, wire together”. Hebbian plasticity involves two main mechanisms, LTP and LTD. LTP is a process by which synaptic connections become stronger (Bliss and Lomo, 1973), conversely, synapses become weaker during LTD (Bear and Malenka, 1994).

Although synaptic plasticity might occur at all excitatory synapses in the mammalian brain, it is most comprehensively studied in the hippocampus.

Synaptic plasticity is widely assumed to provide the neural basis for learning and memory (Langille and Brown, 2018). It is believed that memory is encoded as alterations in the synaptic strength of synapses. Synaptic plasticity can occur on both the presynaptic and the postsynaptic side of synapses (Citri and Malenka, 2008, Ho et al., 2011). Change of synaptic efficacy could be due to presynaptic release of more or less transmitter than previously, or the postsynaptic cell changing its response to the same amount of transmitter, or both may coincide. However, it is widely accepted that synaptic plasticity is predominantly expressed through changes in the number, location, and properties of the postsynaptic receptors (Henley and Wilkinson, 2013).

(40)

Homeostatic plasticity and metaplasticity are non-Hebbian forms of plasticity.

Metaplasticity is often referred to as plasticity of plasticity. More specifically, metaplasticity changes the physiological or the biochemical state of neurons or synapses based on the history of previous synaptic activation, which in turn alters their ability to generate synaptic plasticity (Abraham, 2008). Homeostatic plasticity is the ability of neurons to regulate their excitability relative to network activity in order to stabilize the activity of a neuron or neuronal circuit (Citri and Malenka, 2008).

LTP and LTD

Even though both Cajal and Hebb proposed the idea of synaptic plasticity, this theory was difficult to validate until Terje Lømo and Tim Bliss reported the discovery of LTP in the hippocampus in 1973 (Bliss and Lomo, 1973). They revealed that a brief, high frequency stimulation into the axons of presynaptic neurons constituting the perforant pathway, facilitated a robust, long-lasting increase, or potentiation, in synaptic strength of the postsynaptic neuron in the dentate area of the hippocampus.

At the molecular level, the NMDA receptor-dependent LTP is the best-known form of LTP, in which the release of glutamate from the presynaptic terminal activates both AMPA and NMDA receptors in the postsynaptic membrane. In contrast to AMPA receptors, the pore of the NMDA receptor channel is blocked by magnesium ions, rendering it inactive (Collingridge, 2003). The magnesium ions can only be released from NMDA receptors when the postsynaptic cell is sufficiently depolarized. When the inner membrane is sufficiently depolarized by sodium ion influx through the AMPA channel, magnesium is electrostatically forced away from the NMDA receptor. The opening of NMDA receptors ultimately allows large numbers of calcium ions into the cell. Calcium acts as a second messenger and activates several intracellular signalling cascades such as the protein kinase C and calcium calmodulin-dependent protein kinase II (CaMKII) (Luscher and Malenka, 2012).

(41)

The signalling cascades underlying the induction of LTP are complex. Importantly, activation of the kinases leads to phosphorylation of a number of proteins including AMPA receptors that increase their effectiveness through enhanced ionic conductance of their channels and delivery of AMPA receptors to the synapse (Luscher and Malenka, 2012). LTP can be divided into early phase, which is present 1-2 hours after potentiation and a late phase lasting more than 2 hours (Citri and Malenka, 2008). Insertion of new AMPA receptors into the postsynaptic plasma membrane is believed to be essential for LTP expression in the early phases of LTP, whereas gene transcription and new protein synthesis are required for the late phase LTP. Induction of LTP also triggers enlargement of the spines. Larger spines are associated with larger PSD and greater glutamate-induced current and calcium influx. Further, LTP causes formation and stabilization of new spines (Lai and Ip, 2013).

LTD is induced by prolonged low-frequency stimulation, leading to modest depolarization of the postsynaptic cell. Consequently, the magnesium ion block of the NMDA receptor is less effectively removed. Nevertheless, a moderate increase in postsynaptic calcium concentration is still achieved. Like LTP, the intracellular signalling cascades underlying LTD are complex. In short, calcium influx during LTD activates signalling cascades such as activation of the protein phosphatases calcineurin and protein phosphatase 1. This triggers the internalization of AMPA receptors, which in turn reduces the strength of synapses (Luscher and Malenka, 2012). In contrast to the growth of dendritic spines in response to LTP, a reduction of synaptic strength during LTD leads to shrinkage and retraction of spines (Lai and Ip, 2013).

(42)

Figure 7: LTP and LTD

A simplified model for mechanisms underlying LTP and LTD. A) LTP: Glutamate binds to NMDA and AMPA receptors. NMDA receptors open if the postsynaptic cell is sufficiently depolarized by sodium ion influx through AMPA receptors. Calcium entry through the NMDA channels into the spine activates kinase-dependent cascades, which result in the insertion of additional AMPA receptors into the postsynaptic membrane. B) LTD: Low-frequency stimulation of NMDA receptors cause a modest raise in calcium in the postsynaptic neuron, which initiates phosphatase-dependent cascades, leading to endocytosis of AMPA receptors. Figure adapted from (Dale Purves, 2004).

Synaptic plasticity in disease

Disturbed molecular processes within synapses may potentially lead to neurological and psychiatric disorders. One of the most common neurological disorders is epilepsy, which affects about 1% of the world's population (Fiest et al., 2017). Epilepsy is characterised by a long-lasting and unpredictable predisposition to produce epileptic seizures. A seizure is defined as the excessive and synchronised firing of a population of neurons that last for a short time (Fisher et al., 2005). The excessive firing of neurons will cause increased release of glutamate, which in turn leads to pathological over-stimulation of glutamate receptors. The same mechanism that usually serves to mediate cell to cell signalling may

(43)

during such over-stimulation activate biochemical cascades that promote cell death. This phenomenon is an example of excitotoxicity.

Temporal lobe epilepsy (TLE) is the most common form of epilepsy with focal seizures.

In particular, the hippocampus in the temporal network is associated with TLE (Zhao et al., 2014). Neuroplastic processes have been considered both as a cause and consequence of epilepsy (Jarero-Basulto et al., 2018). Seizure activity can cause widespread induction of LTP and thereby reducing overall hippocampal plasticity available for memory processing (Reid and Stewart, 1997). During TLE, there is selective neuronal cell loss in the CA1 and CA3 regions of the hippocampus, decrement in GABAergic interneurons, reactive gliosis and axonal sprouting in the granular cell layer of the dentate gyrus.

Increased numbers of aberrant synaptic connections together with reduced inhibitory interneurons contribute to the establishment of hyperexcitable circuitry (Jarero-Basulto et al., 2018).

Among the most prevalent mental disorders in the world is major depressive disorder (MDD). The lifetime prevalence of experiencing at least one major depressive episode is 18% (Malhi and Mann, 2018). MDD is characterised by a chronic feeling of sadness or worthlessness, irritability, insomnia, physical lethargy and lack of interest in outside stimuli. There is evidence of structural changes in the brain during depression (Kaltenboeck and Harmer, 2018). Some of these changes can be reversed by treatment with anti-depressants (D'Sa and Duman, 2002). Chronic stress results in reduced dendritic branching, length and spine density in the hippocampus. Moreover, it has been shown that LTP is impaired in rats subjected to stress during induction of a depressive state (Vollmayr and Gass, 2013). Instead, in the fully developed rodent model of depression, LTP is strengthened (Ryan et al., 2010).

Altered mechanisms of synaptic plasticity could lead to neuropsychiatric disorders like epilepsy and MDD. However, the molecular correlates of the plastic changes underlying these conditions are not clearly understood.

(44)

AIMS AND HYPOTHESIS

The overall aim of this thesis has been to increase insight in the molecular mechanisms underlying synaptic plasticity. In particular, the thesis has aimed to determine the ultrastructural localization of the SNARE proteins and synaptotagmin 1 in the postsynaptic compartment, and explore their association with glutamate receptor trafficking within the postsynaptic spine.

Overall hypothesis

1: SNARE proteins and synaptotagmin 1 are expressed in postsynaptic spines.

2: SNARE proteins are co-localized with AMPA and NMDA receptors in the postsynaptic spine.

3: VAMP2 is involved in the exocytotic vesicular insertion of AMPA receptors into the postsynaptic plasma membrane.

4: The postsynaptic concentration of SNAP-25 is increased during LTP.

5: The postsynaptic concentrations of syntaxin-1 and synaptotagmin 1 are changed in animal models of neuropsychiatric disorders.

Aims of the individual articles

Article I: Postsynaptic VAMP/synaptobrevin facilitates differential vesicle trafficking of GluA1 and GluA2 AMPA receptor subunits

The aim of this study was to determine the subsynaptic localization and concentration of VAMP2 at excitatory synapses in the hippocampus. In particular, I wanted to determine whether VAMP2 is expressed in postsynaptic spines. I further aimed to examine co- localization of VAMP2 with the AMPA receptor subunits GluA1 and GluA2, and if so, whether VAMP2 plays a role in the postsynaptic plasma membrane insertion of these subunits.

(45)

Article II: SNARE protein syntaxin-1 colocalizes closely with NMDA receptor subunit NR2B in postsynaptic spines in the hippocampus

The aim of this study was to determine the subsynaptic localization and concentration of syntaxin-1 at excitatory synapses in the hippocampus. Specifically, I wanted to determine whether syntaxin-1 is present in postsynaptic spines. Moreover, I aimed to investigate whether syntaxin-1 is co-localized with the AMPA receptor subunit GluA2 and NMDA receptor subunit GluN2B in the postsynaptic spine.

Article III: A possible postsynaptic role for SNAP-25 in hippocampal synapses

The aim of this study was to determine the subsynaptic localization and concentration of SNAP-25 at excitatory synapses in the hippocampus. Especially, I wanted to determine whether SNAP-25 is expressed in postsynaptic spines. Furthermore, I aimed to examine whether the synaptic concentration of SNAP-25 is changed during the early phase of LTP.

Article IV: The calcium sensor synaptotagmin 1 is expressed and regulated in hippocampal postsynaptic spines

The aim of this study was to determine the subsynaptic localization and concentration of synaptotagmin 1 at excitatory synapses in the hippocampus. Primarily, I wanted to determine whether synaptotagmin 1 is localized in postsynaptic spines. Assuming that synaptotagmin 1 is expressed postsynaptically, I wanted to explore whether synaptotagmin 1 is regulated in postsynaptic spines by using a rat model of chronic TLE.

(46)

Article V: Antibodies raised against aldehyde-fixed antigens improve sensitivity for postembedding electron microscopy

The aim of this study was to compare antibodies raised against antigen pre-fixed with glutaraldehyde to similar commercially available polyclonal and monoclonal standard antibodies. I aimed to evaluate the performance of the antibodies by using common methods like western blot (WB), light microscopy, fluorescence microscopy and electron microscopy. In addition, I wanted to determine whether antibodies raised against glutaraldehyde treated antigen have a stronger affinity for proteins fixed with glutaraldehyde compared to formaldehyde-fixed proteins.

Article VI: Presynaptic PICK1 facilitates trafficking of AMPA- receptors between active zone and synaptic vesicle pool

The aim of this study was to determine the role of PICK1 and NSF in the trafficking of GluA2-containing AMPA receptors at the synapse. In particular, I wanted to investigate whether overexpression of PICK1 or inhibiting NSF-GluA2 interaction with pep2m peptide in hippocampal cultures decreases the concentration of GluA2 in the different compartments of the synapse. Moreover, I wanted to determine the subsynaptic localization and concentration of GluA2-containing AMPA receptors.

Article VII: Changes in concentrations of NMDA receptor subunit GluN2B, Arc and syntaxin-1 in dorsal hippocampus Schaffer collateral synapses in a rat learned helplessness model of depression

The aim of this study was to determine changes in concentrations of selected synaptic proteins in the rat learned helplessness model of depression. By performing a WB pilot experiment on whole-brain synaptosomes, I wanted to identify candidates of synaptic proteins that may show concentration differences between the congenitally learned helpless (LH) compared to congenitally non-learned helpless (NLH) rats. Additionally, I aimed to document any changes of the identified proteins at the subsynaptic level in the hippocampus by using more elaborate quantitative immunogold electron microscopic analyses.

(47)

METHODS

This chapter will, in short, describe the major methods used in this work. Details regarding the methodology used in each study are described in the material and method section of the respective papers.

Antibodies

Antibodies are proteins in the immune globulin family, which specifically recognize their antigens. Antibodies are Y-shaped proteins composed of two heavy chain and two light chains. At the tip of each arm of the Y, there is an antigen-binding site between the light and the heavy chain, which has a unique structure that allows it to bind antigen in a highly specific manner (Saper, 2009). This region is termed as the Fab fragment of the antibody, while the base of the Y is called the Fc region. The Fc portion is responsible for biological activity of the antibody.

Antibodies, particularly for use in immunohistochemistry, represent one of the most powerful tools in biological science. However, the fixation and embedding routine in preparation of tissue for postembedding electron microscopy makes antibody-antigen interactions inefficient. The fixation of tissue with glutaraldehyde for electron microscopical studies denatures the proteins, distorts the target epitopes and influences antigenicity (Stradleigh and Ishida, 2015). The binding efficiency between the antibodies and the epitopes is consequently reduced. To overcome these challenges, we immunized rabbits with SNARE protein antigens that had been fixed with glutaraldehyde. The resulting antisera were then affinity-purified against antigen pre-fixed with glutaraldehyde. Thus, the resulting antibodies would hopefully recognize and bind to the antigens that have been conformationally changed and denatured by glutaraldehyde, the way they are in fixed tissue. A similar approach has been tried previously (Bock et al., 1997, Danbolt et al., 1992).

in synaptic plasticity