Neuropeptide secretion principles Persoon, Claudia Marjolein
2021
document version
Publisher's PDF, also known as Version of record
Link to publication in VU Research Portal
citation for published version (APA)
Persoon, C. M. (2021). Neuropeptide secretion principles: Vesicle populations, characteristics and fusion mechanisms. s.n.
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
E-mail address:
vuresearchportal.ub@vu.nl
Claudia M. Persoon
Neuropeptide secretion principles
Vesicle populations, characteristics and fusion mechanisms
Claudia M. Persoon
OPEPTIDE SECRETION PRINCIPLES
Neuropeptide secretion principles Vesicle populations, characteristics
and fusion mechanisms
Claudia M. Persoon
Neuropeptide secretion principles
Vesicle populations, characteristics and fusion mechanisms Claudia M. Persoon
ISBN/EAN: 978-94-6416-262-2 Copyright © 2020 Claudia Persoon
All rights reserved. No part of this thesis may be reproduced, stored or transmitted in any way or by any means without the prior permission of the author, or when applicable, of the publishers of the scientific papers.
Layout and design by Birgit Vredenburg, persoonlijkproefschrift.nl Printing: Ridderprint | www.ridderprint.nl
Printing of this thesis was financially supported by the CNCR
NEUROPEPTIDE SECRETION PRINCIPLES Vesicle populations, characteristics and fusion mechanisms
ACADEMISCH PROEFSCHRIFT
ter verkrijging van de graad Doctor of Philosophy aan de Vrije Universiteit Amsterdam,
op gezag van de rector magnificus prof.dr. V. Subramaniam, in het openbaar te verdedigen ten overstaan van de promotiecommissie
van de Faculteit der Geneeskunde op vrijdag 23 april 2021 om 13.45 uur
in de aula van de universiteit, De Boelelaan 1105
door
Claudia Marjolein Persoon geboren te Leiden
Copromoter: dr. R.F.G. Toonen
Prof. dr. Susanne Schoch McGover Prof. dr. J. Peter H. Burbach Prof. dr. Anna S. Akhmanova Prof. dr. August B. Smit
Chapter 1 Introduction 9
Chapter 2 Pool size estimations for dense-core vesicles in mammalian CNS neurons
61
Chapter 3 The RAB3-RIM pathway is essential for the release of neuromodulators
107
Chapter 4 Immunocapture of dense-core vesicles using selective transmembrane proteins
179
Chapter 5 General summary and discussion 227
Appendices English summary
Nederlandse samenvatting Publication list
Acknowledgements
252 254 257 258
1
Introduction
INTRODUCTION
The brain is a highly complex organ that regulates our movements, learning and memory, cognition, behavior, emotions and the function of many basic processes in our body. A highly organized network of communicating neurons and glial cells execute these different functions. To communicate, cells release signaling molecules which interact with receptors on receiving cells, thereby triggering intracellular responses. Neurons are the primary cell types responsible for most brain functions. They communicate via two regulated secretion pathways, coordinating fast and slow communication.
Fast communication between neurons occurs at synapses, specialized contact sites where small synaptic vesicles (SVs) release neurotransmitters at a high temporal resolution. Slower communication is regulated by the secretion of neuropeptides, neurotrophins and other signaling molecules, together referred to as neuromodulators. Neuromodulators regulate processes such as development, circadian rhythm, metabolism, plasticity, behavior and emotions. Neurons store and transport neuromodulators in dense-core vesicles (DCVs). DCV release, like SV exocytosis, is triggered by calcium influx. While the mechanism of SV exocytosis has been extensively studied, providing a molecular framework of proteins organizing the release of neurotransmitters, much less is known about neuromodulator secretion.
This thesis provides new insights into the mechanism of regulated secretion of neuromodulators by DCVs.
Neuromodulator signaling in the brain
Neurons and neuroendocrine cell secrete (a) neuropeptides, (b) neurotrophic factors, (c) guidance cues (d) proteases and (e) other vesicular content such as granins and receptors via the regulated fusion of DCVs. These different signaling molecules function in a wide variety of biological processes. Once released they signal through interaction with receptors (Chini et al., 2017).
(a) Neuropeptides are small (2-80 amino acids) neuronal signaling molecules which comprise a diverse group of peptides functioning as neurotransmitters, neuronal modulators or hormones. Currently 48 families, containing more than 200 neuropeptides, have been characterized in mammals (Wang et al., 2015). Most neuropeptides act by binding to G protein-coupled receptors (GPCRs), a large superfamily of proteins differentially expressed in numerous cell types (Wang et al., 2015). Binding to GPCRs induces a conformational
change in the receptor which leads to activation of intracellular G proteins by an exchange of GDP for GTP. G proteins then dissociate and activate a wide range of intracellular signaling pathways (van den Pol, 2012). A large single-cell RNA-sequence dataset of 22,439 mouse cortex neurons showed that 97% of the neurons express one or more neuropeptides and neuropeptide-selective GPCRs (Smith et al., 2019).
18 neuropeptide-precursor genes and 29 neuropeptide-GPCRs (Table 1) were very highly expressed and found in similar areas of the cortex (Smith et al., 2019). Although it was long thought that neuropeptides, once secreted, diffuse over long distances in the extracellular space (Ludwig and Leng, 2006), the co-expression of neuropeptides and their GPCRs implies very local signaling networks exist in the cortex. This is in line with calculations of diffusing neuropeptides released from a single DCV, showing effective concentrations only within a radius of
~55 µm (Chini et al., 2017). How neuropeptides regulate behavior and neuronal functioning through local signaling is largely unknown.
(b) Neurotrophins regulate neuronal development, survival, function and synaptic plasticity in the central nervous system (CNS). They include the brain-derived neurotrophic factor (BDNF), nerve growth factor (NGF), neurotrophin-3 (NT-3) and neurotrophin-4/5 (NT-4/5) (Lessmann et al., 2003). BDNF is the most abundant neurotrophin in the mammalian brain and has been implicated in mechanisms of neuronal functioning and synaptic plasticity (Chapleau et al., 2009; Dean et al., 2009; Lu and Chow, 1999). Neurotrophins signal via two types of receptors, a family of high-affinity tyrosine receptor kinases (Trk) and low-affinity pan- neurotrophin receptors (p75). Receptor binding results in the activation of many signaling pathways, including the Ras/ERK pathway, activating multiple downstream targets including MAP kinases, which promote cell differentiation; phosphatidylinositol-3-kinase/Akt kinase pathway involved in cell survival; and the phospholipase C-y1 pathway, resulting in a signaling cascade that increases release of calcium from internal stores and activates various enzymes (Huang and Reichardt, 2001; Poo, 2001).
1
ble 1. Most abundant expressed neuropeptides from different neuropeptide families and their selective GPCRs in mouse cortex (Smith et al., 2019). ain known functions of neuropeptides are listed. Less expressed neuropeptides (e.g. vasopressin/oxytocin, granins) are not represented in the table. References (Koob and Volkow, 2016) (Baldwin et al., 2010; Edkins, 1906) (Boehm and Betz, 1997; Moore et al., 1988) (Malva et al., 2012) (Merali et al., 1999) (Henning and Sawmiller, 2001) (Hupalo et al., 2019) (Helke et al., 1990) (St.-Gelais et al., 2006) (Bathgate et al., 2003; Kumar et al., 2017) (Leary and Connor, 1995; Wysolmerski, 2012) Function Nociception, stress, reward Digestion, Feeding Modulation (inhibition) of neurotransmission/secretion, cell proliferation Neuroprotection, neurogenesis, communication between the immune- and nervous system Feeding behavior, motor function Vasodilation cognition and behavior during stress stress response, vasodilation, mood behavior, anxiety modulation of dopamine, neurotransmitter and hormone secretion pregnancy and childbirth, collagen turnover, behavior, stress response, feeding proliferation and differentiation endocryne activity, motor function, feeding behavior
Cognate neuropeptide-GPCRs Mu-Opioid Receptor Delta-Opioid Receptor, Kappa- Opioid ReceptorOpioid Receptor- Like 1 Cholecystokinin B Receptor Somatostatin Receptor 1,2,3, 4 Neuropeptide Y Receptor Y1,2,5 GRP Receptor Neuromedin B Receptor VIP Receptor 1,2 ADCYAP1 Receptor 1 CRH Receptor 1,2 Tachykinin Receptor 1,3 Neurotensin Receptor 1,2 Relaxin Family Receptor 1,2,3 PTH 1 Receptor TRH Receptor, -2
Neuropeptide Enkephalins Dynorphins Nociceptins Cholecystokinins Somatostatins Cortistatin Neuropeptide Y Gastrin-Releasing Peptide Neuromedin B Vasoactive Intestinal Peptide Adenylate Cyclase-Activating Polypeptides Corticotropin-Releasing Hormone Neurokinin B Substance P, Neurokinin A Neurotensin Relaxin 1 Parathyroid-Hormone-Like Hormone Thyrotropin-Releasing Hormone
europeptide family pioid gene family astrin/cholecystokinin gene mily omastostatin gene family and Y-amide gene famility ombesin-like peptide gene mily lucagon/secretin gene family orticotropin-Releasing ormone -related gene family inin and tensin (Tachykinin) ne family nsins and Kinins lin family o family neuropeptides
(c) Fundamental processes during embryonic and postnatal development of the nervous system include the migration of neuronal precursor cells to their final destination, and the correct guidance of axons to their targets for connectivity (Guan and Rao, 2003). These processes are steered by extracellular cues, represented by families of guidance cues such as netrins, ephrins, semaphorins and slits. These guidance cues either diffuse or attach to the cell membrane after secretion, creating a complex regulated environment of gradients able to coordinate the development of the nervous system by attracting or repelling axons and neurons (Guan and Rao, 2003).
(d) Proteases are enzymes that hydrolyze peptide bonds between amino acids of a protein, thereby degrading that protein or activating it by breaking inhibitory peptide bonds. Neuropeptides are packed as proproteins in DCVs, and processed by proteases such as proprotein convertases (PCs) to active peptides. The PC family comprises nine members, from which PC1/3 and PC2 are predominantly present in the regulated secretory pathway of the CNS, while furin, PC5/6, PACE4 and PC7 are ubiquitously expressed and process proteins at the trans-golgi network or in the constitutive secretory pathway (Gary, 2007; Seidah et al., 2013; Winsky-Sommerer et al., 2000).
(e) In addition to the different signaling molecules and proteases, DCVs contain granins, receptors, GPCRs, ion channels, ions, ATP and high levels of calcium. The granin family consists of chromogranins and secretogranins; acid proteins with a high capacity to bind calcium.
Chromogranins and secretogranins regulate vesicle sorting, packaging and condensation of soluble cargo and calcium storage inside DCVs (Machado et al., 2010; Yoo and Hur, 2012). Calcium levels can be regulated through interactions between granins and inositol 1,4,5-triphosphate (IP3)-receptor/calcium-channels present in the DCV membrane (Yoo, 2000). In addition to secreting soluble signaling molecules, DCVs also deliver receptors and ion channels to the cell surface. Although multiple proteomics studies have been performed, there is currently no consensus on the DCV proteome (Bark et al., 2012;
Hinckelmann et al., 2016; Wegrzyn et al., 2010; Zhao et al., 2011) and the full extent of the DCV secretome remains to be characterized.
1
Neuropeptides are differentially expressed and their specific expression pattern is a strong identifier for the different neuronal subclasses present in the brain (Fig. 1A; Smith et al., 2019). Different glutamatergic neurons express for example BDNF or Thyrotropin Releasing Hormone (Trh), while clear segregations between NPY, dopamine (Th-expressing neurons), vasoactive intestinal polypeptide (VIP) or Somatostatin (Sst) expressing GABAergic neuronal populations exist (Fig. 1A; Smith et al., 2019; Tasic et al., 2018).
Although some neuromodulators are mainly expressed in specific neuronal subtypes, single neurons show remarkable high expression levels of many different neuropeptides and neuropeptide-receptors. For example, when considering expression of the 18 highest expressed neuropeptides and 29 corresponding receptors (table 1) in Sst expressing, GABAergic interneurons in Layer VI of the cerebral cortex, at least 11 neuropeptides and 17 receptors are (highly) expressed (Fig. 1B-C; Sst. L. VI neurons; Tasic et al., 2018).
Glutamatergic neurons in layer VI, which most likely form local networks with these Sst GABAergic interneurons, show similar high expression levels of multiple neuropeptides, and expression of many receptors which could be activated by neuropeptides released by Sst GABAergic interneurons (Fig.
1B-C). These data show a very complex network of multiple neuropeptide signaling pathways between the different neuron types, but also within a single neuron, creating the possibility of many self-activating feedback loops.
Within a single cell, these multiple neuromodulators are secreted through DCVs, which concentrate large amounts of peptides. Neurosecretory granules, with an approximate diameter of 150-160 nm are estimated to contain 85,000 molecules (Chini et al., 2017; Morris, 1976; Nordmann and Morris, 1984).
Neuronal DCVs are smaller, on average 65-70 nm in diameter (van de Bospoort et al., 2012), which, taken into account the volume difference, would result in an estimation of 8,000 – 10,000 molecules per DCV. Co-storage of multiple neuropeptides/neurotrophins in a single DCV has been shown in multiple CNS neurons (For review: Merighi, 2002). EM immunogold labeling of calcitonin gene-related peptide (CGRP), BDNF and Substance P in axonal projections of the amygdala showed single labeled (~18%), double-labeled (~20%), triple-labeled (not quantified) or non-labeled (~40%; Salio et al., 2007) DCVs, showing heterogeneous loading of DCVs.
1
Figure 1. Neuropeptide and neuropeptide-receptor expression in the brain
(A) T-SNE plot showing neuronal sub-type clusters of single neurons from mouse cortex (primary visual cortex and anterior lateral motor cortex) based on their single-cell RNA sequencing expression profile (https://portal.brain-map.org/atlases-and-data/rnaseq; Tasic et al. 2018). Expression of neuropeptides/neurotrophins in different cell types is visualized.
Specific neuropeptide are expressed in subpopulations of glutamatergic neurons (blue background) or GABAergic neurons (red background) and neuron types can be identified based on their neuropeptide expression profile (Smith et al., 2019). (B) Expression pattern of 18 neuropeptide and 29 neuropeptide-receptor genes with highest expression profile in the brain (Smith et al., 2019) in neuronal subset defined as Sst expressing GABAergic or glutamatergic neurons from Layer VI. (C) Schematic representation of complex neuropeptide signaling network present between Sst-expressing GABAergic and glutamatergic layer VI neurons. Only highest expressed neuropeptides/neuropeptide receptors of measured subset (B) are visualized. Multiple neuropeptides are expressed in single neurons and can be present in a single DCV (Merighi, 2002). Secreted neuropeptides signal through GPCRs, which are highly expressed in the same cell and nearby cells, creating local, complex neuropeptide signaling networks.
The presence of multiple neuromodulators in a single vesicle (Fig. 1C) adds another layer of complexity to the neuromodulator signaling pathway; if there are three or more neuromodulators packed into 1 DCV and they are all secreted, do they activate multiple signaling cascades simultaneously in receiving neurons? For example CGRP, BDNF and Substance P receptors show overlap in intracellular signaling pathways, including the activation of PKC/IP3, protein kinase A or ERK pathways (Reviews: Garcia-Recio and Gascón, 2015; Patapoutian and Reichardt, 2001; Russell et al., 2014). How does a neuron secrete specific neuromodulator signals required at a certain moment/location? Possible mechanisms could include selective secretion of neuromodulators by the modulation of the fusion pore (Chiang et al., 2014) or different dissociation rates from the precipitated vesicle core (Montesinos et al., 2008; Schroeder et al., 1996).
Together, these data indicate that a complex regulated neuromodulator network of multiple neuromodulators per neuron and per vesicle exists, in combination with many receiving receptors at neighboring cells but also at the secreting cell itself. The regulation of this process is largely unknown and many outstanding questions remain unanswered. Undoubtedly a massive amount of energy is directed towards establishing this neuromodulator signaling network, but it is unclear how a neuron provides specificity to neuromodulator signaling. Perhaps different sub-populations of DCVs are present, categorized by different maturation states, release-ready states or signaling proteins present on the vesicle. Different release mechanisms could
operate to regulate differential secretion. Basic insights into the total number of DCVs per neuron, their protein content, maturation states or their release machinery are lacking, which are essential to unravel the functions of this highly organized and vital neuromodulator signaling network.
Neuromodulator signaling in disease
The importance of neuromodulator signaling pathways is highlighted by their essential functions in the brain and body and their association with many disorders, including psychiatric and cognitive disorders, chronic pain, obesity and diabetes (Hökfelt et al., 2018; Malva et al., 2012; Meyer-Lindenberg et al., 2011; Morales-Medina et al., 2010; Vähätalo et al., 2015; Wiesenfeld-Hallin et al., 2002; Yoshimura et al., 2015). Currently, over 30% of clinically marketed drugs target neuromodulator related GPCRs (Wise et al., 2002), indicating that modulations of these secretion pathways provide great potentials for new therapeutic targets. A few examples of implications of neuromodulators in disorders and potential therapeutic targets are listed below.
Complex social behaviors as attachment, recognition, aggression, anxiety and fear conditioning are regulated by neuropeptides such as Oxytocin and Vasopressin. Their receptors are important drug development targets for mental disorders with social deficits such as autism spectrum disorder, social anxiety disorder, borderline personality disorder or schizophrenia (for a review: Meyer-Lindenberg et al., 2011).
Many neuromodulators are implicated in drug-abuse, for example Neurotensin (St.-Gelais et al., 2006), Corticotropin-releasing factor, which orchestrates the stress response and has been implicated in drug-seeking behavior and stress-induced relapse (Valentino and Aston-Jones, 2010; Wise and Morales, 2010) or BDNF, for which enhanced levels correlate with increased drug- seeking behavior (McGinty et al., 2010).
Neuropeptide levels can be dramatically altered during inflammation, injury or pain. Cholecystokinin (CCK) expression is upregulated upon nerve injury, which has negative effects on pain treatment because CCK inhibits the effects of opioids as morphine (Wiesenfeld-Hallin et al., 2002). CGRP levels are elevated in migraine, implicating CGRP as a key player in the pathogenesis of migraine, and clinical trials with CGRP receptor antagonists show effective
1
Many neuropeptides function in feeding behavior and dysregulation has been strongly correlated with pathogenesis of obesity or anorexia. Ablation of Agouti-related peptide (AGRP) and NPY-expressing or proopiomelanocortin (POMC)-expressing hypothalamic neurons in adult mice through cell-type specific application of toxin resulted in a significant reduction in food intake and body weight in mice with AGRP ablated neurons, while POMC-neuron depleted mice showed a gradual increase in body weight and food intake (Gropp et al., 2005). Furthermore, an anorectic mouse line with an autosomal recessive mutation (anx/anx) showed loss of NPY and AGRP in nerve terminal projections and evidence of neurodegeneration of NPY/AGRP expressing neurons (Hökfelt et al., 2008). Transgenic mice overexpressing NPY are obese and show impaired glucose metabolism (Vähätalo et al., 2015). In humans a leucine7 to proline7 polymorphism in preproNPY has been associated with increased diabetes and cardiovascular disease risk (Aton et al., 2005;
Jaakkola et al., 2009; Karvonen et al., 1998). Expression of NPY containing this polymorphism resulted in increased peptide synthesis in AtT-20 cells and cultured hippocampal neuron and enhanced secretion in chromaffin cells, suggesting this polymorphism might affect translation of NPY (Mitchell et al., 2008). Together, these data show that NPY, AGRP and POMC are important regulators of energy homeostasis and disrupting these pathways result in pathogenic effects. Other examples of peptides involved in feeding behaviour include bombesin-related peptides like gastrin-releasing peptide or neuromedin B, which are secreted in response to ingested food and regulate feeding behavior by inhibiting further food intake. They might play important roles in disorders with disrupted food intake behavior, as anorexia, bulimia, obesity and depression (Merali et al., 1999). Treatments for obesity that target bombesin receptors are currently being developed (Chobanian et al., 2012).
These examples show that neuromodulators are essential regulators of physiological functioning of our brain and body, and that alterations in their secreted levels could result in pathophysiological situations. Many treatments target neuromodulator GPCRs, but controlling the levels of secreted neuromodulators would also be a very effective treatment option for many disorders. To develop new therapeutic targets and progress clinical brain research, insights into the neuronal pathway of neuromodulator secretion by DCVs is very important.
Life of a DCV
Identification of DCVs in neurons
Progress in the field of biochemistry and the development of the electron microscope in the 1950s and 1960s made it possible to characterize different cellular components. Multiple types of vesicles are found in cells, participating in for example constitutive or regulated secretory pathways or the endo- lysosomal pathway (box 1). Regulated secretion of neuromodulators occurs through DCVs. DCVs are recognized on electron micrographs as dense-cored vesicles with a diameter of 70-100 nm. In 1954, Nobel prize laureate George Palade together with Sanford Palay were one of the first to describe the presence of these vesicles in nerve terminals (Palade and Palay, 1954). They noted that these “dense bodies, presumably lipid inclusions, are scattered throughout the cytoplasm”. Furthermore, Palade was the first to describe dense vesicles present near the endoplasmic reticulum in pancreatic cells and linked these granules and the ER to secretory pathways (Palade, 1956).
Following decades of research focused mainly on describing the presence and behavior of DCVs (e.g. Pellegrino De Iraldi & Etcheverry 1967), and pioneering work in the field of neuroendocrinology in the 1960s and 1970s was directed towards isolating and characterizing different neuropeptides secreted from the brain (Hökfelt, 1991; Zupanc, 1996). Later studies focused on the biogenesis, transport and fusion of DCVs (van de Bospoort et al., 2012;
Kim et al., 2006; McMahon et al., 1992; Renden et al., 2001; Silverman et al., 2005; Verhage et al., 1991; de Wit et al., 2006). Although many fundamental questions remain, a general description of the different steps in the neuronal DCV pathway is emerging and will be discussed here.
DCV biogenesis
Cargos of DCVs are synthesized in the endoplasmic reticulum (ER) as proproteins and transported to the cis-Golgi network (Fig. 2B). As these proproteins move through the Golgi stacks and trans-Golgi network, posttranslational modifications occur until they are packed into secretory vesicles. Proteins are sorted to different secretory pathways (box 1) depending on sequence motifs and protein domains mainly present in the pro-domain of the protein. Proteins destined for the regulated secretory pathway are packed highly concentrated into DCVs. Multiple proteins have been implicated in this step of DCV biogenesis, including chromogranins, secretogranins, carboxypeptidase E, sortilin or HID-1 (Gondré-Lewis et al.,
1
2003), but how proteins are correctly sorted to the constitutive or regulated secretory pathway, possibly via active sorting or retention mechanisms, is poorly understood.
DCV Maturation
DCVs exiting from the trans-Golgi are generally considered “immature” and undergo maturation steps, including acidification of the granular lumen, processing of peptides and possibly sorting of proteins by membrane fission, which may also be proceeded by homotypic fusion, see below. An acidic pH and high calcium concentration in the vesicle are required for aggregation of DCV cargo and for processing of prohormones by prohormone convertases.
This intra-vesicular environment is achieved and regulated by V-ATPase proton-pumps (Kim et al., 2006). Although necessary for sorting, the pro- domain of a prohormone or neuropeptide needs to be cleaved of for the peptide to mature. Cleavage of the pro-domain occurs in the (trans-) Golgi network, in the vesicle lumen or extracellular, depending on the pH-optimum of the required enzyme (Lessmann and Brigadski, 2009).
Homotypic fusion, a fusion event between similar vesicles, is thought to mature DCVs by sorting different cargo and membrane proteins and increasing the size of immature DCVs (Du et al., 2016; Klumperman et al., 1998; Tooze et al., 1991; Urbé et al., 1998). This process has only been established in cell- free assays, and it remains a question whether it occurs in neuronal cells.
Modern technology provides the opportunity to study DCVs at single vesicle resolution, and will make it possible to follow-up on this phenomenon of homotypic fusion and establish whether this mechanism is involved in DCV maturation in neurons.
Transport of DCVs
DCVs are transported throughout the neuron by microtubule-based motor proteins of the kinesin and dynein protein families (Fig. 2C; Lo et al., 2011;
Zahn et al., 2004). In the axon, microtubules are orientated uniformly with their plus-ends out, resulting in anterograde transport driven by kinesin motors and retrograde transport directed by dynein motors (Fig. 2C). Dendrites contain microtubules of mixed polarity, resulting in bidirectional movement of motor proteins (Fig. 2C; Kapitein and Hoogenraad, 2011)
1
Figure 2. Biogenesis, transport and exocytosis of neuromodulator containing DCVs.
(A) Schematic representation of a mammalian hippocampal neuron. (B) DCV biogenesis in the soma (area indicated in A). Cargos of DCVs are synthesized in the endoplasmic reticulum (ER) as pro-proteins and transported to the Golgi network. Neuromodulator pro- proteins and transmembrane (TM) proteins move through the Golgi stacks and trans-Golgi network until they are packed into secretory vesicles. (C) DCVs are transported throughout the neuron (area indicated in A) by microtubule-based motor proteins of the kinesin and dynein protein families. Microtubules are orientated uniformly with their plus-ends out in the axon, resulting in anterograde transport driven by kinesin motors and retrograde transport directed by dynein motors. Microtubule polarity is mixed in dendrites. DCVs can contain more than one motor protein. (D) DCV exocytosis occurs at synaptic and extra-synaptic regions (area indicated in C) and is triggered by calcium influx. DCVs can fuse completely with the plasma membrane, releasing all cargo (Full fusion), or release part of their cargo and reseal (kiss-and-run).
Studies in primary cultures of rat neurons and in C. elegans identified KIF1A as a motor protein driving anterograde DCV transport in the axon (Lo et al., 2011; Zahn et al., 2004) and retrograde transport by dynein is coordinated through interaction with dynactin, as shown in Drosophila (Lloyd et al., 2012).
Directionality and selective transport of DCVs is regulated by recruitment or activation of specific motor proteins to the vesicle. This motor-cargo attachment process is regulated by multiple proteins such as, vesicle membrane receptors, integral membrane proteins, signaling proteins, scaffolding proteins or small RAB GTPases with help of RAB effector molecules (Kapitein and Hoogenraad, 2011; Karcher et al., 2002; Schlager and Hoogenraad, 2009; Lloyd et al., 2012).
Whether DCVs detach from the microtubule network and how they arrive at a fusion site is currently unknown. At the Drosophila neuromuscular junction, an activity-evoked increase in capture of DCVs at nerve terminals has been reported, providing a mechanism for activity dependent capture of DCVs at fusion sites (Shakiryanova et al., 2006). In mammalian neurons, DCVs show an arrest in transport upon calcium influx (de Wit et al., 2006), but whether this represents capture at fusion sites is currently unknown. Furthermore, the exact localization and requirements for a DCV fusion site is not known. On average 60% of the DCV fusion events occur at synapses (van de Bospoort et al., 2012; Farina et al., 2015). Electron micrograph data shows that a synapse at rest contains approximately 1 DCV (van de Bospoort et al., 2012). But, while SVs are predominantly localized to the active zone, the area where they fuse at the pre-synapse, DCVs are typically located further away from
the active zone (Verhage et al., 1991; Zupanc, 1996). Whether DCVs fuse at similar areas as SVs at the synapse is an interesting question, as it might provide insights into shared release requirements between the two types of secretory pathways.
Trigger for DCV exocytosis
The major trigger for regulated exocytosis is an increase in calcium concentration by entry of external calcium through voltage-gated calcium channels and calcium release from intracellular calcium stores such as the ER. Upon an action potential, the membrane depolarizes and calcium levels are elevated by influx of extracellular calcium through voltage-gated calcium channels. One action potential raises calcium levels high enough for SVs to fuse, but DCVs require prolonged depolarization and bulk calcium concentration increase to trigger release (Leenders et al., 1999; Verhage et al., 1991). SVs are closely localized to voltage-gated calcium channels, therefore local calcium concentration increase are sufficient to trigger their release (Bucurenciu et al., 2008; Holderith et al., 2012). DCVs in contrast, are predominantly located outside the active zone and therefore not in close proximity of voltage-gated calcium channel. This could explain the requirement of prolonged depolarization before the onset of DCV release;
global calcium levels need to rise within the cytoplasm of the cell before the calcium concentration is high enough in proximity of the vesicle.
Duration and extent of calcium increase can be modified by (i) mitochondria, which buffer calcium concentrations through uptake of the ion, (ii) by the distribution of calcium channels and stores, and (iii) by other calcium regulating pathways. In neuroendocrine cells, the strength of stimulus- secretion coupling is influenced by the spatial distribution of all these factors in relation to the vesicle, which influences the exocytosis rates of DCVs (review: Kits and Mansvelder, 2000). Furthermore, following influx from calcium channels, intracellular stores can release their calcium. DCV mobility and exocytosis in Drosophila neuromuscular junction (NMJ) nerve terminals is influenced by calcium-induced calcium release from ER through the downstream effector Calmodulin kinase II (CamKII; Shakiryanova et al., 2007), showing that calcium-induced calcium release enhances neuromodulator secretion, at least in fly NMJ.
1
DCV fusion
Upon a calcium trigger, DCVs fuse with the plasma-membrane to release their content (Fig. 2D). This fusion step is regulated by the soluble N-ethylmaleimide sensitive factor attachment protein receptor (SNARE) complex and supporting proteins (see section molecular mechanism of regulating secretion). DCV fusion events show high variation in the amount of cargo they release, ranging from fusion events that result in complete release of their cargo to short fusion events with minimal cargo release (de Wit et al., 2009; Xia et al., 2009). After release of their entire cargo, DCVs can dissolve by incorporating into the plasma membrane (full fusion or full-collapse; Fig. 2D). In case of incomplete cargo release, the fusion pore reseals and the vesicle can fuse again at later time-points (kiss-and-run; Fig. 2D). Parameters introducing variability in fusion events include fusion time, amount of fusion pore expansion and place of fusion (Chiang et al., 2014; de Wit et al., 2009; Xia et al., 2009). These factors could be cargo dependent and introduce a layer of plasticity in the regulation of DCVs cargo release (Burgoyne and Morgan, 2003).
Box 1. Constitutive and regulated secretory routes in the neuron
All cells require a secretory pathway to replenish material at the plasma membrane and to communicate with other cells. The secretion of proteins, membrane components and other substances is mainly accomplished by the constitutive secretory pathway. In addition, specialized secretory cells contain a regulated secretory pathway for fast, on demand secretion of neurotransmitters (by SVs), neuromodulators, hormones or enzymes (by secretory granules or DCVs). At the trans Golgi network, newly synthesized proteins, originating from the ER, are packed into vesicles and directed to a secretory pathway (Fig. A). Vesicles destined for constitutive secretion are transported to the plasma membrane and fuse, releasing their content or incorporating their membrane components.
Secretory granules or DCVs are transported throughout the cell but require a regulated trigger before fusion with the plasma membrane occurs (Fig. A; section Life of a DCV). In addition to the regulated secretion of neuromodulators, neurons release neurotransmitters by SVs at highly specialized pre-synaptic terminals (Fig. A; see Synaptic vesicle exocytosis section below). Some specialized neurons exist with diverging secretory routes as for example dopamine neurons (See dopamine release section
below). In addition to the secretion of proteins, material is also retrieved from the plasma membrane and extra-cellular space through endocytosis.
Material present in the endo-lysosomal pathway can either be recycled via the different secretory pathways or degraded via lysosomes (Fig. A;
Burgoyne and Morgan, 2003; Kim et al., 2006; Südhof and Rizo, 2011).
Synaptic vesicle exocytosis
At pre-synaptic nerve terminals, small synaptic vesicles (~40 nm diameter) release neurotransmitters with high speed and spatial precision. A presynaptic nerve terminal contains hundreds of SVs in different states of fusion-readiness; a small percentage of SVs attach/dock to the plasma membrane at the active zone and can be primed, making them ready for immediate fusion upon arrival of an action potential. In addition, a large reserve pool of vesicles resides at the synapse which are released after prolonged stimulation. Upon an action potential, calcium flows into the nerve terminal via calcium channels and exocytosis is triggered within less than a millisecond. The release sites of SVs are precisely opposed to postsynaptic receptors of the connecting neuron, resulting in fast neuronal communication.
After fusion, vesicles are endocytosed, recycled and refilled with neurotransmitters, creating a local recycling pool of SVs at synapses which does not rely on replenishment of vesicles from the Golgi (Rizzoli and Betz, 2005; Südhof and Rizo, 2011).
Monoamine release
Monoamine neurotransmitters, as dopamine, noradrenaline, serotonin or histamine, are a class of neurotransmitters functioning in many physiological functions as emotions, learning and memory. They are secreted through different types of vesicles. Dopamine release mechanisms have been studied in detail. Dopamine is secreted from dopamine neurons located in the mid-brain. It signals through GPCRs and regulates motor function, motivation and reward. Dopamine vesicles are small, clear vesicles, resembling SVs. But, in contrast to classic neurotransmitters, dopamine is secreted from axons, dendrites and the soma. Dopamine axons form varicosities filled with vesicles that are usually not associated with a post-synaptic density. The varicosities contain clusters of active-zone proteins RIM, ELKS2 and Bassoon,
1
Figure A. Neuronal secretory pathways. Schematic representation of the different secretory pathways in neurons.
Cargos for secretory vesicles are synthesized in the endoplasmic reticulum (ER) and transported to the Golgi network where they are packed into vesicles destined for the constitutive secretory pathway or regulated secretory pathway. Vesicles in the constitutive pathway (purple) are transported and immediately fuse with the plasma membrane, while vesicles in the regulated pathway (red, DCVs) require a trigger before fusion occurs. SVs (orange) secrete neurotransmitters at synapses in a fast and spatial restricted manner.
The pool of SVs is locally recycled via endocytosis. The endo-lysosomal pathway (pink) regulates recycling and degradation.
Action potentials trigger two different modes of release; a mild, widespread tonic release triggered by non-synchronous spontaneous activity, or fast and transient phasic release due to synchronized burst firing. Fast dopamine release occurs within milliseconds of an action potential and has a high release probability (Liu et al., 2018), setting it apart from neuromodulator release (van de Bospoort et al., 2012). After release, dopamine is thought to diffuse in tissue to activate relatively distant receptors, a process named volume transmission (Reviews: Grace, 2016; Liu and Kaeser, 2019).
Molecular mechanism of regulated secretion
The process of exocytosis is supported by a universal machinery for membrane fusion which is highly conserved across species. Much of our knowledge of the secretory pathway is based on initial identification of essential secretory proteins in yeast by Peter Novick and Randy Schekman.
By combining the then latest developments in genetics, biochemistry and electron microscopy, they developed a screen to select temperature-sensitive Saccharomyces cerevisiae mutants with secretion deficits based on buoyant density, which was increased due to the accumulation of proteins and membranes (Novick and Schekman, 1979). Using this screen, they initially discovered 23 SEC proteins which drive the secretory pathway in yeast (Novick and Schekman, 1979; Novick et al., 1980, 1981). Ten of these act in the final steps of secretion (SEC1-6, 8-10 and 15; Novick et al., 1981), together with the later identified SNC1/2 (Protopopov et al., 1993) and SSO1/2 (Aalto et al., 1993). Subsequent research in C. elegans, identifying UNC proteins essential for secretion (named after the uncoordinated movement phenotype of the worms), Drosophila, and mammalian neurons revealed that orthologs for many of the initially identified proteins in yeast drive regulated secretion of vesicles (See box 3 for different model systems; Jahn and Scheller, 2006;
Südhof, 2013; Südhof and Rothman, 2009). These include two groups of highly conserved proteins; the soluble N-ethylmaleimide sensitive factor attachment protein receptor (SNARE) complex, containing SEC9 (SNAP25; J et al., 1993;
Sollner et al., 1993), SNC1/2 (VAMP/synaptobrevin; Schoch et al., 2001), SSO1/2 (Syntaxin; Bennett et al., 1992), and SM (SEC1/MUNC18-like; Verhage et al., 2000) proteins. In contrast, orthologs of other SEC proteins, involved in final secretion steps, are largely dispensable for SV fusion, including the GTPase SEC4 and its effectors: SEC2, a guanine exchange factor for SEC4 (Walch-Solimena et al., 1997), and 6 subunits of the SEC4 effector complex called the exocyst complex, SEC3, 5, 6, 8, 10 and 15 (Bowser et al., 1992;
Guo et al., 1999; Mehta et al., 2005; Murthy et al., 2003; Schlüter et al., 2004;
TerBush and Novick, 1995; TerBush et al., 1996). The role of these orthologs in regulated secretion in mammalian neurons remains poorly understood.
Our current knowledge concerning the involvement of the different SNARE proteins and interacting proteins in regulated secretion in neurons derives mainly from studies on fast neurotransmitter release (Figure 3 + Box 2 for detailed information on fusion mechanism of SV exocytosis). Here I discuss
1
the different proteins that have a known function in DCV exocytosis and focus on MUNC13, RIM1/2 and RAB3 proteins, which will be studied in Chapter 3.
SNARE, SM proteins and calcium sensor
Of the different SNARE proteins, Synaptobrevin I, -II/VAMP and SNAP25 have been shown to be required for DCV fusion. Synaptobrevin II/VAMP is required for both SV and DCV fusion in CNS neurons (Schoch et al., 2001;
Shimojo et al., 2015), while Synaptobrevin I regulates SV fusion in a subset of hippocampal neurons, in line with its expression profile (Schoch et al., 2001). Sensory neurons of the trigeminal ganglia require Synaptobrevin I for the release of CGRP, a neuropeptide which mediates dilation of intracranial blood vessels (Meng et al., 2007). These data suggest that both SV and DCV fusion is regulated by different Synaptobrevin/VAMP isoforms, depending on the expression of the v-SNAREs in different cell types. SNAP-25 is required for DCV fusion in hippocampal neurons (Arora et al., 2017; Shimojo et al., 2015), and the SNAP homologs SNAP-23 and SNAP-29, but not SNAP-47, support DCV fusion to different extent upon overexpression in SNAP-25 KO neurons (Arora et al., 2017).
The role of MUNC18 proteins in neuropeptide secretion from CNS neurons has not been addressed, but chromaffin cells derived from Munc18-1 null mice show a 10-fold reduction in DCV fusion and a 10-fold reduction in the number of docked DCVs (Voets et al., 2001). Furthermore, peptide hormone secretion in the pituitary gland is almost completely inhibited (Korteweg et al., 2005), suggesting an important role for MUNC18-1 in docking and fusion of DCVs.
Most members of the synaptotagmin protein family, encoded by 17 mammalian genes, are thought to function as calcium sensor for regulated exocytosis.
Different paralogs have been reported to be present on DCVs, including synaptotagmin-1, the major calcium sensor involved in SV exocytosis (Walch- Solimena et al., 1993) and synaptotagmin-4 (Bharat et al., 2017; Dean et al., 2009). Synaptotagmin-IV (syt-IV) knock out posterior pituitary neurons have less exocytosis of DCVs upon stimulation (Zhang et al., 2009), while BDNF release is increased in hippocampal neurons (Dean et al., 2009). Also synaptotagmin-10 was found to act as a calcium sensor for DCVs secreting Insulin-like growth factor 1 (IGF-1) in the olfactory bulb (Cao et al., 2011), but expression levels are very low in the CNS, making it an unlikely candidate as calcium sensor for DCVs in the CNS. Which syntaptotagmin(s) functions as
Figure 3. SV release machinery
Schematic representation of the release machinery for synaptic vesicles. SNARE and SM proteins (right) together with priming and active zone proteins (left) regulate fusion of synaptic vesicles at the pre-synaptic active zone. See box 2 for functional details of
1
Box 2. Proteins regulating SV exocytosis (related to figure 3)
SNARE complex
SNARE proteins are characterized by the presence of a SNARE motif and are required for most intracellular membrane trafficking processes. The neuronal SNARE complex mediating SV fusion consists of the proteins synaptobrevin-2/VAMP (vesicle-associated membrane protein), present on the vesicle (R-SNARE), SNAP-25 (synaptosomal-associated protein of 25 kDA) and Syntaxin-1, both present on the target plasma membrane (Q-SNAREs). During fusion, the SNARE complex assembles into a tight complex by interaction of the SNARE motifs, forming a parallel four-alpha- helix bundle, bringing the vesicular and plasma membrane into close proximity. The energy released upon assembly of the SNARE complex results into the fusion of the two membranes, opening the fusion pore and allowing the cargo of a vesicle to be released (Brunger et al., 2018;
Burgoyne and Morgan, 2003; Kasai et al., 2012; Rizo and Xu, 2015;
Südhof and Rizo, 2011). Dissociation of the SNARE complex after fusion is accomplished by NSF and SNAPs (soluble NSF-attachment proteins;
Burgoyne and Morgan, 2003; Südhof and Rizo, 2011).
SM proteins
Mammalian SM proteins are essential partners in the assembly of the SNARE complex to accomplish fusion and regulators of final trafficking steps before fusion. The various SM proteins are evolutionary conserved cytosolic proteins that fold into an arch-shaped structure and have a high binding affinity for their specific syntaxin, which contains a conserved region available for interaction (Südhof and Rizo, 2011; Toonen and Verhage, 2003).
MUNC18-1 proteins are required for SV fusion, as spontaneous and evoked release of SVs is completely abolished in Munc18-1 knockout mice (Verhage et al., 2000). MUNC18-1 binds to the N-terminal end of syntaxin-1, and, together with MUNC13, provides a functional template that guides the assembly of a SNARE complex with SNAP25 and Synaptobrevin/VAMP, thereby stabilizing release-ready vesicles (He et al., 2017; Jiao et al., 2018; Meijer et al., 2012; Südhof and Rizo, 2011;
Toonen and Verhage, 2003; Wang et al., 2019).
