has been associated with fear and anxiety pathways (Koch et al., 2000; Wojcik et al., 2013).
Characteristic of all members of the UNC-13/Munc13 protein family is the highly conserved C-terminal region which contains two Munc13 homology domains (MHD) connected by a linker (Koch et al., 2000). This region, referred to as the MUN domain, has been shown to weakly interact with the N-terminus of Syntaxin-1 (Betz et al., 1997;
Ma et al., 2011, 2013). It has been shown that the MUN domain is sufficient for the basic priming function of Munc13s since it can at least partially rescue the Munc13-deficient phenotype in neurons and chromaffin cells (Basu et al., 2005; Stevens et al., 2005). The binding of Munc18-1 to the N-terminus of Syntaxin-1 renders Syntaxin-1 in a closed conformation, which led to the hypothesis that the priming function performed by Munc13s is based on their ability to interact with Syntaxin-1. Munc13s could therefore induce a switch from the closed to an open Syntaxin-1 conformation, which would enable and/or accelerate SNARE complex nucleation (Ma et al., 2011, 2013; Sassa et al., 1999).
The C-terminus of Munc13 proteins possesses a single C2 (C2C) domain, the function of which remains unknown (Brose et al., 1995; Shin et al., 2010). Further towards the N-terminus, the MUN domain is preceded by accessory domains that execute important regulatory roles for Munc13 function. Adjacent to the MUN domain, Munc13s contain a second C2 domain (C2B) which is the only C2 domain of UNC-13/Munc13 proteins to bind phospholipids in a Ca2+-dependent manner, with a preference for phosphoinositides like phosphatidylinositol phosphate (PIP) and phosphatidylinositol 4,5-bisphosphat (PIP2) (Shin et al., 2010). The C2B domain is preceded by a C1 domain that is homologous to the diaglycerol (DAG)/phorbolester-binding domain of protein kinase C. Point mutations that disable the binding of DAG to Munc13-1 cause perinatal lethality in mice and a significant decrease in the RRP without changes in the EPSC amplitude, resulting in an increase in the vesicular release probability (Pvr) in autaptic neuron cultures (Ahmed et al., 1992;
Aravamudan et al., 1999; Lackner, 1999; Maruyama and Brenner, 1991; Rhee et al., 2002). Preceeding the C1 domain, members of the UNC-13/Munc13 proteins contain a Calmodulin-binding region, which is functionally highly conserved and mediates Ca2+ -dependent Calmodulin-binding to UNC-13/Munc13 proteins (Aravamudan et al., 1999; Hu et al., 2013; Junge et al., 2004; Lipstein et al., 2012, 2013; Rodríguez-Castañeda et al., 2010). Mice which express a Munc13-1 variant that has lost its ability to bind Ca2+ -Calmodulin, are viable, develop normally and show no major changes in basic synaptic transmission characteristics. However, these mice exhibited a deficit in synaptic vesicle priming during high activity at the calyx of Held synapse, a highly specialized
glutamatergic synapse in the auditory pathway that is an established model to study plasticity characteristics (Lipstein et al., 2013).
In contrast to the highly conserved C-termini of UNC-13/Munc13 proteins, the N-terminal domain structures vary significantly between isoforms. The only exceptions are UNC-13L in C.elegans and Munc13-1 and ubMunc13-2 in mammals, which exhibit homologous N-terminal sequences. These isoforms contain a third C2 domain (C2A), which does not bind phospholipids in a Ca2+-dependent manner like the C2B domain, but mediates binding to the Rab3a-interacting molecule (RIM), a cytoskeletal component of the presynaptic active zone that has been shown to have a role in localizing Ca2+-channels to the active zone and in docking and priming synaptic vesicles in their vicinity (Andrews-Zwilling et al., 2006; Betz et al., 2001; Fernández-Busnadiego et al., 2013; Han et al., 2011; Kaeser et al., 2011, 2012).
Mice lacking Munc13-1 die perinatally and electrophysiological recordings from glutamatergic hippocampal autaptic neurons in culture revealed a severe decrease in the RRP size measured by hyperosmotic sucrose solution, in the spontaneous release of synaptic vesicles and in Ca2+-dependent evoked release (Augustin et al., 1999a).
Augustin et al., observed no changes in the number of synaptic vesicles and in the number of docked vesicles in electron micrographs after chemical fixation (Augustin et al., 1999a). Mice deficient of both Munc13-1 and Munc13-2 isoforms die prenatally and exhibit a complete shutdown of excitatory and inhibitory neurotransmission in hippocampal neuron cultures with no apparent changes in synapse density, synapse morphology, synaptic vesicle density and the number of membrane-attached synaptic vesicles in electron micrographs after chemical fixation (Varoqueaux et al., 2002).
Taken together, it is evident that members of the Munc13 family are essential proteins in neurotransmission since null mutants cause a complete loss of spontaneous and evoked synaptic transmission. The fact that application of hypertonic sucrose solution fails to trigger vesicle release in these mutants implies a complete absence of readily-releasable fusion-competent vesicles, despite normal numbers of morphologically docked synaptic vesicles in electron micrographs from chemically fixed samples. These findings established Munc13s as proteins important for mediating a molecular priming step downstream of synaptic vesicle docking. However, synapses lacking UNC-13 in C.
elegans and Munc13-1 and -2 in mice were recently shown to exhibit an almost complete loss in the number of docked synaptic vesicles, analyzed using improved cryo-fixation
12
Weimer et al., 2006). These findings indicated that Munc13s, major molecules in synaptic vesicle priming, also have a role in synaptic vesicle membrane-attachment and that the concept of subsequent docking and priming steps prior to synaptic vesicle fusion has to be revised.1.3.3. Ca2+-dependent activator protein for secretion (CAPS)
The Ca2+-dependent activator protein for secretion (CAPS) family consists of two ~145 kDa proteins, which were originally identified as essential cytosolic factors for regulated Ca2+-dependent release in PC12 cells (Walent et al., 1992). CAPS proteins are highly conserved across species and have structural homology to members of the Munc13 family (Koch et al., 2000). Whereas C. elegans and Drosophila only express one CAPS isoform (UNC-31 & dCAPS), it has been shown that mammals express two isoforms, CAPS-1 and -2 (Jockusch et al., 2007; Speidel et al., 2003). CAPS proteins contain a single Munc13 homology domain, which includes a Syntaxin-1-binding region (Betz et al., 1997; Khodthong et al., 2011; Koch et al., 2000). Moreover, the domain structure of CAPS proteins contains an N-terminal dynactin-binding region, followed by a C2 domain that can bind phospholipids in a Ca2+-dependent manner, indicating a role of CAPS in Ca2+ -dependent membrane interactions (Grishanin et al., 2002; Sadakata et al., 2007a). The C2 domain is followed by a pleckstrin homology (PH) domain, a structural motif that can associate with acidic phospholipids of the plasma membrane and that can bind to PIP2 (Grishanin et al., 2002). A C-terminal stretch of acidic amino acids mediates binding of CAPS to LDCVs and seems crucial for normal CAPS function in PC12 cell secretory vesicle release in addition to the required Ca2+-dependent PIP2 binding to the PH domain (Grishanin et al., 2002, 2004). For many years, most studies had indicated a specific role of CAPS in LDCV priming and release in neuroendocrine tissues (Elhamdani et al., 1999;
Grishanin et al., 2002, 2004; Liu et al., 2008, 2010; Rupnik et al., 2000; Speidel et al., 2003, 2005, 2008; Tandon et al., 1998; Wassenberg and Martin, 2002).
Western blot analysis and immunotainings for CAPS-1 reveals the presence of CAPS-1 in neuroendocrine tissues, for example in chromaffin cells of the adrenal medulla, in glucagon-containing α-cells of the pancreas, and in endocrine cells of the anterior pituitary (Sadakata et al., 2007b; Walent et al., 1992; Wassenberg and Martin, 2002). However, CAPS-1 is also strongly expressed in the brain, especially in the hippocampus and in the cerebellar granule cell layer, where its immunolabelling pattern co-localizes with the synaptic vesicle marker Synaptophysin, indicating a synaptic localization of the protein
(Speidel et al., 2003; Walent et al., 1992; Wassenberg and Martin, 2002). CAPS-2 expression appears to be distinct from CAPS-1 and displays a strong level of co-localization with neutrophin-3 and brain-derived neurotrophic factor (BDNF), both factors important for neurodevelopmental processes in cerebellum (Sadakata et al., 2006, 2007c;
Speidel et al., 2003). The widespread presence of CAPS in synaptic terminals in almost all brain regions, including terminals which do not contain many LDCVs, raised the question as to whether or not CAPS proteins have a role in synaptic vesicle release (Jockusch et al., 2007).
In Drosophila neuromuscular junction synapses, the loss of CAPS causes a severe reduction in LDCV release reflected by an accumulation of LDCVs in the terminals, but showed additional defects in the release of synaptic vesicles, with a reduction of the excitatory responses by 50% and a morphological accumulation of synaptic vesicles (Renden et al., 2001). However, the defects in glutamatergic synaptic vesicle release could not be rescued by genetically targeted expression of rat CAPS at the neuromuscular junction, leading to the interpretation that CAPS might influence fast synaptic transmission indirectly by preventing the delivery of components of the synaptic vesicle release machinery by LDCVs (Renden et al., 2001). Mice lacking both CAPS isoforms do not survive birth, whereas mice lacking only CAPS-2 are phenotypically indistinguishable at birth from wild-type mice (Jockusch et al., 2007; Speidel et al., 2003). Hippocampal autaptic neurons cultures from CAPS-1/2 double knock-out (DKO) mice display a complex phenotype. In 38% of all neurons, no evoked or spontaneous release could be detected, with a non-measurable RRP after hypertonic sucrose application (Jockusch et al., 2007).
The remaining 62% of the cells exhibited a 68% decrease in the EPSC amplitude after stimulation and an 85% reduction in the size of the RRP. Synapse morphology and density as well as the number of total and docked synaptic vesicles in glutamatergic synapses were unchanged in electron micrographs of chemically fixed samples (Jockusch et al., 2007).
The current model poses that CAPS may prime synaptic vesicles by regulating SNARE complex assembly, since CAPS stimulates formation of trans-SNARE complexes from Syntaxin-1/SNAP25 acceptor and Synaptobrevin-2 donor in liposomes fusion assays (Daily et al., 2010; James et al., 2008, 2009, 2010; Khodthong et al., 2011). However, the role of CAPS proteins in synaptic vesicle priming is still heavily debated as many groups claim that the observed physiological deficits in neurotransmission in CAPS-1/2 DKO mice