Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Mol Cell Neurosci. Author manuscript; available in PMC 2014 January 1.
Published in final edited form as:
PMCID: PMC3540146

Differential regulation of evoked and spontaneous neurotransmitter release by C-terminal modifications of Complexin


Complexins are small α-helical proteins that modulate neurotransmitter release by binding to SNARE complexes during synaptic vesicle exocytosis. They have been found to function as fusion clamps to inhibit spontaneous synaptic vesicle fusion in the absence of Ca2+, while also promoting evoked neurotransmitter release following an action potential. Complexins consist of an N-terminal domain and accessory α-helix that regulate the activating and inhibitory properties of the protein, respectively, and a central α-helix that binds the SNARE complex and is essential for both functions. In addition, Complexins contain a largely unstructured C-terminal domain whose role in synaptic vesicle cycling is poorly defined. Here, we demonstrate that the C-terminus of Drosophila Complexin (DmCpx) regulates localization to synapses and that alternative splicing of the C-terminus can differentially regulate spontaneous and evoked neurotransmitter release. Characterization of the single DmCpx gene by mRNA analysis revealed expression of two alternatively expressed isoforms, DmCpx7A and DmCpx7B, which encode proteins with different C-termini that contain or lack a membrane tethering prenylation domain. The predominant isoform, DmCpx7A, is further modified by RNA editing within this C-terminal region. Functional analysis of the splice isoforms showed that both are similarly localized to synaptic boutons at larval neuromuscular junctions, but have differential effects on the regulation of evoked and spontaneous fusion. These data indicate that the C-terminus of Drosophila Complexin regulates both spontaneous and evoked release though separate mechanisms and that alternative splicing generates isoforms with distinct effects on the two major modes of synaptic vesicle fusion at synapses.

Keywords: Exocytosis, Synapse, Neurotransmitter Release, Synaptic Vesicle, SNARE complex


A conserved vesicle trafficking machinery made up of soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) and SM proteins drives membrane fusion in multiple cellular compartments (Südhof and Rothman, 2009). At synapses, the fusion of neurotransmitter containing vesicles with the plasma membrane is tightly regulated to allow for precise communication within the nervous system. Several SNARE-binding accessory proteins that modulate the activity of the core machinery have evolved to provide synapse-specific requirements for synaptic vesicle fusion. The best studied of these accessory proteins is the vesicular Ca2+ sensor synaptotagmin 1 (Syt 1), which binds to SNARE complexes and membrane phospholipids in a Ca2+-dependent manner to allow for fast synchronous neurotransmitter release in response to Ca2+ (Geppert et al., 1994; Yoshihara and Littleton, 2002; Xu et al., 2007). In contrast to Syt 1, the precise function of Complexin (Cpx) in synaptic vesicle fusion is still being elucidated. Cpxs are small, α-helical proteins identified based on their ability to bind the assembled SNARE complex with 1:1 stoichiometry (McMahon et al., 1995; Pabst et al., 2000; Bracher et al., 2002; Chen et al., 2002).

Numerous studies have suggested that Cpx acts both to inhibit spontaneous neurotransmitter release in the absence of Ca2+ (Huntwork and Littleton, 2007; Maximov et al., 2009; Hobson et al., 2011; Martin et al., 2011) and to promote evoked neurotransmitter release (Reim et al., 2001; Huntwork and Littleton, 2007; Xue et al., 2007, 2008; Cai et al., 2008; Maximov et al., 2009; Hobson et al., 2011; Martin et al., 2011). Data from biochemical studies (Giraudo et al., 2006; Schaub et al., 2006), genetic knock-out studies in Drosophila and C. elegans (Huntwork and Littleton, 2007; Hobson et al., 2011; Martin et al., 2011) and genetic knock-down studies in mice (Maximov et al., 2009) have supported the role of Cpx as an inhibitor of spontaneous neurotransmitter release. Genetic deletion of the single Cpx homolog in Drosophila (DmCpx) results in a dramatic increase in the frequency of spontaneous vesicle fusion events (minis) at the larval neuromuscular junction (NMJ) (Huntwork and Littleton, 2007; Cho et al., 2010). Similarly, the frequency of tonic fusion events at the C. elegans NMJ is increased in genetic knock-outs of the primary Cpx homolog (CeCpx-1) (Hobson et al., 2011; Martin et al., 2011). Unlike flies and worms, mammals have four Cpx genes with distinct expression patterns in the nervous system (Reim et al., 2005). RNAi knock-down of Cpxs in mouse cortical cultures increases spontaneous neurotransmitter release (Maximov et al., 2009). However, genetic knock-out of Cpxs results in decreased spontaneous neurotransmitter release at hippocampal autapses and GABA-/glycinergic synapses, but not at striatal autapses (Xue et al., 2007, 2008; Strenzke et al., 2009).

In contrast to the different findings on spontaneous fusion, studies have consistently shown that Cpx is necessary to promote evoked Ca2+-dependent neurotransmitter release. These data indicate that Cpx has distinct effects on different modes of neurotransmitter release and plays several roles during the multi-step process of synaptic vesicle fusion. Structure-function studies suggest that different domains of Cpx contribute to specific steps in synaptic vesicle trafficking. A central helix within Cpx is necessary for SNARE binding as determined by crystallography (Bracher et al., 2002; Chen et al., 2002). Cpx constructs that lack this domain or key binding residues within it are non-functional (Xue et al., 2007; Giraudo et al., 2008; Maximov et al., 2009; Cho et al., 2010; Martin et al., 2011). The N-terminus, meanwhile, appears to contain both facilitatory and inhibitory domains that may be differently used at mammalian and invertebrate synapses (Xue et al., 2007, 2009; Giraudo et al., 2009; Xue et al., 2010; Hobson et al., 2011; Martin et al., 2011). In contrast, the function of the C-terminus is poorly understood. Biochemical studies have shown that the C-terminus inhibits SNARE-mediated cell fusion but promotes cell-mediated liposome fusion (Giraudo et al., 2008; Malsam et al., 2009). In addition, Cpx constructs that lack the C-terminus are functional in hippocampal autapses, yet fail to rescue the increased tonic neurotransmitter release observed at the C. elegans NMJ in cpx-1 null mutants, suggesting the C-terminus may act to inhibit neurotransmitter release at some synapses. Recent studies of several mammalian Cpx isoforms suggest the C-terminal domain may differentially regulate clamping versus activation properties of different isoforms (Kaiser-Woo et al., 2012). Given these divergent results, additional characterization of the C-terminus is needed to define its precise role in synaptic transmission.

In this study, we analyzed the function of the C-terminus of DmCpx. Using a chemical mutagenesis approach, we isolated a Cpx allele with an early stop codon that truncates the far Cterminus. These mutants show reduced Cpx protein levels and mislocalize Cpx at synaptic boutons at Drosophila larval NMJs. We subsequently identified two alternatively spliced isoforms, DmCpx7A and DmCpx7B, which vary in the far C-terminus, with additional C-terminal variation created through RNA editing of DmCpx7A. Although DmCpx7A predominates at the mRNA level in larvae and adults, both isoforms are expressed in the developing nervous system and their mRNA expression is activity-regulated. In transgenic rescue experiments, we show that DmCpx7A and DmCpx7B are similarly localized to synaptic boutons at the larval NMJ. However, we find that DmCpx7A and DmCpx7B have different effects on spontaneous and evoked neurotransmitter release, with DmCpx7A being a better inhibitor of spontaneous release and DmCpx7B functioning as a better facilitator of evoked release. DmCpx7A contains a C-terminal membrane tethering prenylation domain, while DmCpx7B does not. We propose a model in which C-terminal modification regulates the effects of Cpx on different modes of neurotransmitter release.


The C-terminus of DmCpx is necessary for protein stability, localization and function

We performed an ethane methylsulfonate (EMS) non-complementation mutagenesis screen with the cpxSH1 null mutant to isolate additional Cpx alleles in Drosophila. cpx572 was identified in a screen of 5,000 mutagenized lines as a loss-of-function mutation. Sequence analysis revealed that the cpx572 allele contains a small deletion near the end of exon 6, leading to a premature stop codon and deletion of the final ~25 residues of the Cpx protein (Fig. 1A). Similar to the cpxSH1 null allele (Huntwork and Littleton, 2007), cpx572 is semi-lethal. Homozygous adult cpx572 escapers are uncoordinated and ataxic, though they are less severely affected than null animals. Western blot analysis indicates that a truncated form of Cpx protein is present in head extracts from cpx572 adults. The truncated Cpx protein is expressed at < 20% of wild type levels (Fig. 1B), suggesting that the far C-terminus regulates Cpx protein stability in vivo.

Fig. 1
The Complexin C-terminus is necessary for protein stability and localization. (A) Diagram showing subdomains of the Cpx protein as first reported byXue et al. (2007). The numbering corresponds to residues in DmCpx7A. The region of the C-terminus deleted ...

We next evaluated Cpx protein localization and function at larval NMJs of cpx572 mutants. Although weak Cpx expression could be detected within motor axons, Cpx failed to accumulate in synaptic boutons at similar levels to the wildtype protein when examined by confocal microscopy (Fig. 1C). Imaging at higher laser power revealed that the remaining Cpx found at mutant cpx572 NMJs was distributed in puncta, rather than being localized diffusely in the bouton as observed in control (Fig. 1D). The Cpx puncta found at cpx572 NMJs resided near, but did not completely overlap active zones, as revealed by co-staining for the active zone protein Bruchpilot. The residual Cpx did not overlap with peri-active zone proteins such as Fas2 (Fig. 1D). These data indicate that the far C-terminus of Cpx regulates its subsynaptic localization. In addition to mislocalization of Cpx at synapses, cpx572 larvae displayed disruptions in synaptic transmission at the NMJ, with a slightly less severe phenotype than observed in cpxSH1 null larvae (Fig. 2). cpx572 mutants show a large elevation in the frequency of spontaneous release (Fig. 2A, B), and a milder reduction in the amplitude of the evoked response compared to nulls (Fig. 2C, D). Compared to cpxSH1 null mutants, cpx572 mutants also showed milder defects in synaptic depression during short (Fig. 2E, F) and long (Fig. 2G-I) stimulation trains. These data indicate that the C-terminus regulates both subsynaptic localization and function of Cpx, and that the cpx572 mutant is a hypomorphic allele.

Fig. 2
Synaptic defects in animals lacking the Complexin C-terminus. (A) Postsynaptic current recordings of spontaneous release at muscle 6 NMJs in control (black), cpx572 (brown) and cpx−/− (cpxSH1 - red). This genotypic color code is maintained ...

The Drosophila Cpx locus generates multiple splice isoforms

To further investigate the function of the C-terminus of DmCpx, we used cDNA analysis to detect any endogenous sequence variation present at the Cpx genomic locus. Unlike mammals, which have four Cpx genes, there is only a single Cpx gene in Drosophila. An analysis of the genomic sequence compared to sequenced expressed sequence tag (EST) clones revealed multiple alternative splicing predictions for DmCpx (Fig. 3A). The genomic locus is predicted to encode 12 different transcripts produced by alternative splicing of the first exon, which contains a large portion of the 5’ untranslated region (UTR) of the cpx mRNA, but no coding sequence. Alternative splicing of the first exon is well supported by sequences of ESTs that correspond to those found in the alternate first exons. In general, the 5’ UTR of mRNA sequences can regulate both translation initiation and mRNA stability (Pickering and Willis 2005). The wide variety of 5’ UTRs in cpx transcripts suggests that the timing, level, and/or location of Cpx expression is likely to be tightly regulated in Drosophila. We also identified alternative splicing that altered coding exons within Cpx. Notably, the far C-terminus is encoded by one of two alternatives for exon 7, which we term exon 7A and exon 7B. We PCR amplified and sequenced individual cDNAs from the ATG start site in exon 3 to the stop codon in exon 7A or exon 7B. Although alternative splicing was predicted for exon 4, only exon 4B (Fig. 3A) was ever found from a total of 70 sequenced cDNAs, and alternative splicing of exon 5 resulted in the insertion of only a single amino acid for transcripts utilizing exon 5A compared to exon 5B. By far the most extensive sequence variation occurred due to alternative splicing of exons 7A and 7B, encoding the final 24 and 20 residues, respectively, in the far C-terminus of Cpx. Hereafter, the most abundant transcripts utilizing exon 7A and exon 7B will be referred to as DmCpx7A and DmCpx7B, respectively.

Fig. 3
The Drosophila Cpx locus generates multiple splice isoforms. (A) Scale diagram indicating alternative splice variants of Cpx. Protein-coding regions are shaded. The genomic deletion in cpx572 mutants and Pumilio binding sites in exon 8A are indicated. ...

The C-terminus of Cpx is not highly conserved from invertebrates to mammals, particularly in the far C-terminus encoded by exon 7 in DmCpx (Fig. 3B). However, DmCpx7A contains a C-terminal CAAX-box, similar to mammalian Cpx 3 (mCpx3) and mCpx4, which allows for post-translational prenylation by the addition of either a farnesyl (15-carbon chain) or geranylgeranyl (20-carbon chain) moiety to the cysteine residue and elimination of the terminal three amino acids (Zhang and Casey, 1996). Prenylation can mediate membrane association and protein-protein interactions (Zhang and Casey, 1996), and this CAAX-box is implicated in proper localization of DmCpx7A, mCpx3 and mCpx4 (Reim et al., 2005; Xue et al., 2009; Cho et al., 2010). DmCpx7B lacks a C-terminal CAAX-box, similar to mCpx1 and mCpx2. The unique exon expressed by DmCpx7B is highly conserved among 10 related Drosophila species (data not shown), indicating evolutionary pressure to maintain both isoforms of Cpx. These results suggest that although Drosophila encodes only a single Cpx gene, alternative splicing generates functional variants that contain or lack a CAAX-box, similar to mammalian Cpxs.

To determine expression differences between DmCpx7A and DmCpx7B, we designed primers specific for exon 7A or exon 7B and used quantitative RT-PCR to calculate the relative abundance of DmCpx7A and DmCpx7B transcripts. Among both adult and larval transcripts, DmCpx7A was ~1,000-fold more abundant than DmCpx7B (Fig. 3C), suggesting DmCpx7A is the predominant isoform and that alternative splicing of exon 7 is not developmentally regulated at the time points studied. We also examined whether the relative abundance of DmCpx7A or DmCpx7B was regulated by neuronal activity using a previously reported approach with the temperature-sensitive mutants paralytic (paraTS1) and seizure (seiTS1) (Guan et al., 2005). seits1 is a hyperactivity mutant that disrupts an ERG potassium channel that functions in action potential repolarization. parats1 mutants have reduced neuronal excitability secondary to dysfunction in the voltage-gated sodium channel. Total RNA samples were prepared from paraTS1 and seiTS1 adults following different heat shock protocols to mimic neuronal hypoactivity and hyperactivity, respectively. In the acute heat shock protocol, animals were given a 20-min heat shock at 37°C followed by a 30-min recovery period. In the chronic heat shock protocol, animals were given four 5-min heat shocks at 37°C spaced 1 hour apart and followed by a 24-hour recovery period. At baseline without heat shock, both DmCpx7A and DmCpx7B levels were elevated in seiTS1 mutants (p < 0.01 and p < 0.001, respectively, two-way ANOVA Bonferroni post-test) and DmCpx7B levels were reduced in paraTS1 mutants (p < 0.001) compared to wild-type controls. There was no difference in the activity-dependence of DmCpx7A vs. DmCpx7B (p > 0.05, two-way ANOVA; Fig. 3D). However, both transcripts were significantly downregulated in seiTS1 mutants following the acute or chronic heat shock protocols vs. untreated seiTS1 animals (p < 0.001, two-way ANOVA Bonferroni post-test). In summary, our results suggest neuronal activity levels can modulate Cpx mRNA expression, potentially altering basal release properties of the synapse in an activity-dependent manner.

RNA editing of the DmCpx7A C-terminus generates four unique isoforms

In addition to alternative splicing, DmCpx expression can be regulated post-transcriptionally by RNA editing.Hoopengardner et al. (2003) used a comparative genomics approach with 18 Drosophila species to identify targets of the RNA editing enzyme adenosine deaminase acting on RNA (ADAR) and found three sites of RNA editing within DmCpx exon 7A: position 375, which encodes isoleucine (Ile) residue 125, and positions 388 and 389, which encode asparagine (Asn) residue 130. RNA editing at position 375 changes Ile-125 to methionine (Met), and RNA editing at positions 388 and 389 can change Asn-130 to aspartate (Asp), serine (Ser) or glycine (Gly). The amino acid possibilities at residue 130 are intriguing as Asp is phospho-mimetic, Ser is phospho-competent and Gly is phospho-incompetent, raising the possibility that phosphorylation at this site may regulate DmCpx7A function. The ADAR enzymes require a double-stranded RNA substrate, which is usually provided by an imperfect duplex in the pre-mRNA formed by base pairing between the exon containing the adenosine(s) to be edited and an intronic region called the editing site complementary sequence (ECS) (Higuchi et al., 1993). To investigate the occurrence of RNA editing in DmCpx exon 7A, we compared intronic sequences from 12 Drosophila species to identify highly conserved regions that could function as the ECS. A region of near perfect conservation was found in the final 65 bp of intron 6 extending through the first 88 bp of exon 7A (Fig. 4A). The mfold algorithm for prediction of RNA secondary structure (Zuker, 2003) suggests that this region folds into an extended, imperfect RNA duplex (Fig. 4B), satisfying the requirements for an ADAR substrate.

Fig. 4
RNA editing alters the C-terminus of DmCpx7A. (A) Alignment of an almost perfectly conserved region of Cpx intron 6 and exon 7 among 12 Drosophila species. The three sites of RNA editing in exon 7A are indicated with arrows. (B-C) Predicted RNA secondary ...

To determine the extent of RNA editing of Cpx in Drosophila, we sequenced individual cDNA segments of the editing region amplified by RT-PCR. Interestingly, editing was never observed at the second and third sites (A388, A389) unless editing had also occurred at the first site (A375). According to the mfold algorithm, editing at A375 is predicted to extend the length of the imperfect RNA duplex (Fig. 4C), making it a potentially more stable or attractive substrate for ADAR to carry out editing at sites A388 and A389. At these final two sites, editing was either observed at A389, giving rise to a Ser at residue 130, or at both A388 and A389, giving rise to a Gly at residue 130, but not at A388 alone. Thus, we identified four RNA editing isoforms of DmCpx exon 7A in vivo: Ile125/Asn130 (the unedited isoform), Met125/Asn130, Met125/Ser130, and Met125/Gly130. As a percentage of 96 individual cDNAs sequenced, each isoform was represented as follows: Ile125/Asn130 = 61.5%, Met 125/Asn130 = 7.3%, Met125/Ser130 = 10.4%, and Met125/Gly130 = 20.8%. Only a small subset of transcripts are known to be RNA edited, and nearly all encode proteins involved in rapid neurotransmission (Hoopengardner et al., 2003). Although it is unclear how editing alters Cpx function, mRNA sequencing indicates multiple isoforms of the edited protein are likely to be present at synapses that could differentially impact release properties.

DmCpx7A and DmCpx7B show similar localization and mobility at the larval NMJ

Given that cpx572 mutants showed defects in Cpx localization at synaptic boutons, we sought to determine whether DmCpx7A and DmCpx7B might display differential localization at synapses due to their highly divergent C-termini. We first carried out in situ hybridization analysis of late-stage embryos to determine their expression pattern. Antisense hybridization probes specific for DmCpx7A and DmCpx7B were designed from the 3’ untranslated regions of exon 7A and exon 7B, respectively. Both DmCpx7A and DmCpx7B showed clear expression in the developing central nervous system and ventral nerve cord (Fig. 5A). Consistent with the RTPCR results, the DmCpx7A signal was more robust than that of DmCpx7B. These results suggest that both DmCpx7A and DmCpx7B are expressed in the developing nervous system without an obvious region-specific distribution.

Fig. 5
DmCpx7A and DmCpx7B have similar expression patterns in the developing nervous system and at the NMJ. (A) Lateral view of in situ hybridization on late stage embryos with DmCpx7A and DmCpx7B cRNA probes. Sense cRNA probes are shown as negative controls. ...

We next looked at transgene expression at the larval NMJ. As we were unable to generate antisera specific to DmCpx7A and DmCpx7B, we turned to the Gal4-UAS system (Brand and Perrimon, 1993). We generated transgenic animals expressing DmCpx7A or DmCpx7B mRNAs downstream of a UAS sequence. When driven pan-neuronally with the elavC155-Gal4 driver in the cpx null background (cpxSH1), robust expression of both DmCpx7A and DmCpx7B was observed in synaptic boutons at the larval NMJ by immunostaining with a pan-Cpx antiserum (Huntwork and Littleton, 2007) (Fig. 5B). The halo-shaped distribution of both transgenic proteins was reminiscent of synaptic vesicle-associated proteins such as Syt 1 (Littleton et al., 1993). We next co-expressed DmCpx7A and DmCpx7B transgenic proteins tagged with either GFP or mCherry with the elavC155-Gal4 driver in the cpx null background (cpxSH1). Both proteins showed similar subcellular distribution within presynaptic boutons (Fig. 5C), suggesting either isoform of the protein can target effectively to synaptic terminals.

In transfected neurons, prenylation of mCpx3 and mCpx4 results in a distinct distribution from that of mCpx1 and mCpx2, which lack prenylation (Reim et al., 2005). mCpx3 and mCpx4 appear to be concentrated around synaptic terminals, whereas mCpx1 and mCpx2 are distributed more diffusely throughout axonal processes. Although DmCpx7A, which contains a prenylation motif, and DmCpx7B, which lacks a prenylation motif, showed similar localization at the larval NMJ at the level of confocal microscopy, we wondered whether prenylation of DmCpx7A might result in membrane targeting to synaptic vesicles or the synaptic plasma membrane that would alter its mobility compared to DmCpx7B. We generated N-terminal GFP-tagged DmCpx7A and DmCpx7B transgenic animals and used fluorescence recovery after photobleaching (FRAP) to follow the dynamics of the two Cpx isoforms at synapses. As a control, we used animals expressing UAS-GFP alone to compare to dynamics of soluble proteins, and UAS-Synaptogyrin-GFP to compare to dynamics of synaptic vesicle proteins. We selected large diameter synaptic boutons at muscle fiber 6 and used FRAP to photobleach half of the bouton. We then followed recovery of the GFP-tagged proteins into the bleached area for 1 minute in 10-second intervals using a Perkin Elmer spinning disk confocal microscope. Quantitative analysis of FRAP recovery revealed that GFP-tagged DmCpx7A and DmCpx7B behaved similarly, with a slower diffusion rate than cytosolic GFP alone, but faster than the synaptic vesicle membrane-tethered Synaptogyrin-GFP (Fig. 6A, B). We conclude that both Cpx isoforms are highly dynamic within individual synaptic boutons. If DmCpx7A is associated with synaptic vesicles or the plasma membrane via prenylation, our data indicate it can cycle on and off, resulting in faster recovery dynamics than integral membrane synaptic vesicle proteins. Alternatively, a significant population of Cpx7A might lack prenylation altogether.

Fig. 6
DmCpx7A and DmCpx7B show similar mobility within NMJ boutons. (A) To determine the rate of diffusion of DmCpx7A-GFP and DmCpx7B-GFP within boutons, FRAP was performed on half the diameter of an individual synaptic bouton (boxed area) to assay intra-synaptic ...

DmCpx7A and DmCpx7B differentially regulate spontaneous and evoked neurotransmitter release

To evaluate functional differences between DmCpx7A and DmCpx7B, we tested the ability of the transgenes to rescue the synaptic transmission defects of cpx null mutants (cpxSH1) when expressed pan-neuronally. cpx null mutants alone have a dramatic elevation of spontaneous release and a reduced evoked response (Fig. 2, Fig. 7). DmCpx7A fully rescued the elevated mini frequency of cpxSH1 mutants, whereas DmCpx7B was only partially able to clamp spontaneous fusion (control: 2.5 ± 0.3 Hz, cpxSH1: 80.0 ± 5.0 Hz, DmCpx7A rescue: 3.5 ± 0.3 Hz, DmCpx7B rescue: 10.0 ± 1.0 Hz) (Fig. 7A, B). We compared the ability of DmCpx7A and DmCpx7B to rescue spontaneous release with their putative mammalian (mouse) counterparts that are prenylated like the 7A isoform (mCpx4) or lack a prenylation motif like the 7B isoform (mCpx1). Similar to the effects of the prenylated DmCpx7A isoform, mCpx4 was able to robustly rescue the enhanced spontaneous fusion rate to control levels (Fig. 7A, B). In contrast, the unprenylated mCpx1 isoform failed to fully rescue the enhanced spontaneous release, similar to DmCpx7B. These results indicate a functional diversification of the Drosophila splice variants with respect to clamping properties, similar to what has been achieved with gene duplication in mammals. To further examine differences in DmCpx7A and DmCpx7B, we measured evoked neurotransmitter release in transgenically rescued animals expressing either isoform. DmCpx7A was able to rescue the reduced amplitude of EJCs of cpxSH1 mutants, whereas DmCpx7B showed a strong enhancement of EJC amplitude to more than twice that seen in controls (control: 6.5 ± 0.7 nA, cpxSH1: 4.0 ± 0.3 nA, DmCpx7A rescue: 7.2 ± 0.7 nA, DmCpx7B rescue: 16.0 ± 1.5 nA) (Fig. 7C).

Fig. 7
DmCpx splice variants and orthologous mammalian isoforms share similar clamping properties. (A) Postsynaptic current recordings of spontaneous release at muscle 6 synapses in control (black), cpx−/− null mutants (red), Elav-GAL4;; DmCpx7A, ...

We next compared the rescue of evoked responses by the Drosophila splice isoforms to that achieved with mCpx1 and mCpx4. We observed that mCpx4 again behaved similarly to DmCpx7A, rescuing the evoked defect in cpxSH1 mutants (Fig. 7C). In contrast, mCpx1 behaved like DmCpx7B, enhancing evoked release to far greater levels than controls (Fig. 7C), similar to previous observations (Cho et al., 2010). We conclude that DmCpx7A is a better inhibitor of spontaneous release, while DmCpx7B is a better facilitator of evoked neurotransmitter release. These properties are similar to the rescue effects observed for the prenylated (mCpx4) and unprenylated (mCpx1) mammalian isoforms in our model. A number of reports have suggested that Cpx has different effects on spontaneous vs. evoked neurotransmitter release (Huntwork and Littleton, 2007; Maximov et al., 2009; Cho et al., 2010; Hobson et al., 2011; Martin et al., 2011). Our data indicate that the C-terminus of Drosophila Cpx regulates both spontaneous and evoked fusion, and that alternative splicing generates isoforms with distinct effects on the two release mechanisms.


Although SNAREs make up the core fusion machinery in all cells, additional SNARE-associated proteins are required for rapid Ca2+-mediated neurotransmitter release at synapses. Along with the vesicular Ca2+ sensor Syt 1, Cpx has emerged as a key regulator of SNARE-mediated function. By binding to assembled neuronal SNARE complexes with 1:1 stoichiometry, Cpx is ideally situated to regulate vesicle fusion, but its precise role remains unclear. Initial genetic studies in mice and Drosophila suggested opposing functions in evoked versus spontaneous release (Xue et al., 2008; Huntwork and Littleton, 2007), and in vitro studies of fusion driven by SNAREs and Cpx gave mixed results (Giraudo et al., 2006; Schaub et al., 2006; Malsam et al., 2009; Seiler et al., 2009). Genetic knock-outs of mCpx 1, 2 and 3 resulted in a reduction in both mini frequency and evoked neurotransmitter release in hippocampal autaptic cultures, suggesting a facilitatory role for Cpx in synaptic vesicle fusion. Genetic analysis of the single Cpx gene in Drosophila revealed a dramatic increase (> 40-fold) in mini frequency and a reduction in evoked neurotransmitter release at the larval NMJ, suggesting that Cpx has different effects on these two modes of synaptic vesicle fusion.

More recent studies using RNAi knock-down of Cpxs in mouse cortical cultures (Maximov et al., 2009) and genetic knock-outs of Cpx-1 at the C. elegans NMJ (Hobson et al., 2011; Martin et al., 2011) found similar results to the original Drosophila study (Huntwork and Littleton, 2007), supporting a model in which Cpx inhibits spontaneous vesicle fusion and promotes synchronous evoked vesicle fusion. What distinguishes these two modes of vesicle fusion at the molecular level, however, remains unclear. Structure-function studies of Cpx (Xue et al., 2007, 2009, 2010; Giraudo et al., 2009; Hobson et al., 2011; Martin et al., 2011) have indicated the presence of opposing inhibitory and facilitatory subdomains in the N-terminus of the protein. It has been suggested that the facilitatory and inhibitory functions of the Cpx N-terminus may be weighted differently depending on the model system and synapse examined (Xue et al., 2009), explaining some of the divergent results regarding the effect of Cpx on spontaneous release. The role of the C-terminus, however, is less clear. In vitro, the C-terminus of mCpx1 appears to inhibit SNARE-mediated cell-cell fusion (Giraudo et al., 2008), but promotes SNARE-mediated liposome fusion (Malsam et al., 2009). On the other hand, Cpx constructs lacking the C-terminus are functional in hippocampal autaptic culture (Xue et al., 2007), but fail to rescue the aldicarb sensitivity of cpx-1 null mutants in C. elegans (Martin et al., 2011). Finally, lentiviral rescue of shRNAi mCpx1/mCpx2 knockdown cultured cortical neurons indicated that the Cpx C-terminus regulated clamping and vesicle priming functions of the protein, but not Ca2+-triggering of evoked release (Kaeser-Woo et al., 2012). These results suggest that the C-terminus may exert an effect on neurotransmitter release at some, but not all, synapses.

To address the uncertainty regarding the function of the C-terminus of Cpx, we identified and characterized a Drosophila mutant cpx572, which contains an early stop codon that eliminates the final ~25 residues of the C-terminus. These mutants generated a truncated form of Cpx that was present at ~20% of wild type levels in adult head extracts and mislocalized at synaptic boutons at the larval NMJ. In addition, cpx572 mutants displayed similar, though slightly milder, defects in synaptic transmission compared to cpx null animals. These results argue that the C-terminus of DmCpx is important for protein stability and function. We subsequently examined naturally occurring splice isoforms of DmCpx and identified a splicing event that leads to two alternative far C-termini. These isoforms, DmCpx7A and DmCpx7B, differ in the final ~25 residues of the protein (the same region that is truncated in cpx572 mutants). DmCpx7A is the predominant isoform at the mRNA level in both larvae and adults (Fig. 3C), and both DmCpx7A and DmCpx7B transcripts show similar activity dependence in the temperature-sensitive activity mutants paraTS1 and seiTS1 (Fig. 3D). However, the presence of multiple Pumilio binding sites in the 3’ UTR of DmCpx7A, but not DmCpx7B (Fig. 3A), raises the possibility that the translation of these isoforms may undergo additional regulation. In addition, DmCpx7A undergoes RNA editing within exon 7 to generate further diversity. Editing at one of the sites in DmCpx7A generates a potential phosphorylation site within this domain. Like the edited version of DmCpx7A, the DmCpx7B isoform also contains a potential phosphorylation site in exon 7. Future experiments with edited and non-edited transgenes will be required to define how RNA editing of Cpx regulates its effects on neurotransmission.

Based on the altered Cpx protein localization at the larval NMJ in cpx572 mutants (Fig. 1C), we hypothesized that DmCpx7A and DmCpx7B would show distinct localization patterns at synapses. Although isoform-specific antisera failed to recognize any Cpx immunoreactivity on fixed tissue, we could detect expression of DmCpx7A and DmCpx7B transgenes using pan-Cpx antiserum (Huntwork and Littleton, 2007) or transgenically expressed flouroscently-tagged varients. At the level of confocal microscopy, both transgenes localized to synaptic boutons at the larval NMJ in a halo pattern (Fig. 5B) reminiscent of synaptic vesicle-associated proteins such as Syt 1 (Littleton et al., 1993). These data indicate that although the far C-terminus of Cpx is necessary for proper protein localization, the dissimilar C-terminal sequences of DmCpx7A and DmCpx7B both allow for trafficking to synapses. We also performed FRAP analysis of GFP-tagged DmCpx7A and DmCpx7B GFP-tagged proteins at individual synapses. Both proteins were highly mobile within bleached boutons, recovering faster than a synaptic vesicle integral membrane protein, but slower than cytoplasmic GFP (Fig. 6).

Finally, we examined the function of DmCpx7A and DmCpx7B in synaptic transmission at the larval NMJ by expressing the transgenes in a cpx null mutant background (cpxSH1) and evaluating their ability to rescue the elevated mini frequency and decreased evoked neurotransmitter release observed in cpxSH1 mutants (Huntwork and Littleton, 2007). DmCpx7A was able to restore control levels for both mini frequency and evoked neurotransmitter release (Fig. 7A-B). In contrast, DmCpx7B only partially rescued the elevated mini frequency but promoted evoked neurotransmitter release to more than twice the levels of control (Fig. 7A-B). A number of studies have shown distinct roles for Cpx in spontaneous vs. evoked neurotransmitter release. Besides the increased mini frequency and reduced evoked neurotransmitter release observed in Drosophila cpx mutants, similar results have now been reported at the C. elegans NMJ (Hobson et al., 2011; Martin et al., 2011). In mice, knockouts of mCpx 1, 2 and 3 show decreases in both spontaneous and evoked neurotransmitter release in hippocampal autaptic cultures and certain synapses in brain slices (Xue et al., 2008). In contrast, RNAi knock-down of mCpx 1 and 2 results in increased spontaneous and decreased evoked neurotransmitter release in cortical cultures (Maximov et al., 2009), similar to findings at the Drosophila larval and C. elegans NMJs. Thus, the emerging trend is that Cpx can have distinct effects on different modes of neurotransmitter release. Our findings with DmCpx7A and DmCpx7B indicate the differential ability to inhibit spontaneous neurotransmitter release and promote evoked neurotransmitter release can be regulated by the far C-terminus of Cpx.

The most striking difference between the two far C-terminal DmCpx isoforms is the presence of a prenylation motif in DmCpx7A that is absent from DmCpx7B. In mammals, this motif is found in mCpx 3 and 4, but not in mCpx 1 and 2 (McMahon et al., 1995; Reim et al., 2005), and it has been shown to be important for Cpx protein localization (Reim et al., 2005) and function (Xue et al., 2009; Cho et al., 2010). The prenylation motif in mCpx 3 and 4 is necessary for concentrated expression of these proteins near sites of synaptic vesicle release in transfected neurons, and mutations in this motif result in a more diffuse distribution of mCpx 3 and 4, similar to that seen for mCpx 1 and 2, which lack a prenylation motif (Reim et al., 2005). Furthermore, our lab and others have shown that the prenylation motif in DmCpx7A is critical for its function in neurotransmitter release (Cho et al., 2010).Xue et al. (2009) found that mutating the DmCpx7A prenylation motif impaired its ability to inhibit spontaneous neurotransmitter release in hippocampal autaptic cultures. Similarly,Cho et al. (2010) showed that replacing the prenylation motif of DmCpx7A with the final four residues of mCpx1 not only blocked its ability to inhibit spontaneous neurotransmitter release, but also enhanced its ability to promote evoked neurotransmitter release. In short, disrupting the C-terminal prenylation motif of DmCpx7A caused it to function like DmCpx7B. In addition, the differential effects of rescue on evoked and spontaneous release by DmCpx7A and DmCpx7B was similar to that observed for their mammalian counterparts that were prenylated (mCpx4) or not (mCpx1). These data indicate that the far C-terminus, including C-terminal prenylation at the CAAX box, plays a key role in determining how Cpx clamps spontaneous fusion and promotes evoked release.

Based on our Drosophila Cpx studies, we propose a model for how DmCpx7A and DmCpx7B might differentially regulate neurotransmitter release (Fig. 8). It has been estimated that three to ten SNARE complexes are needed per vesicle to achieve normal Ca2+-triggered vesicle fusion kinetics (Karatekin et al., 2010; Mohrmann et al., 2010). In contrast, only a single SNARE complex is required for in vitro fusion (van den Bogaart et al., 2010), a process that may more directly apply to random spontaneous minis in vivo. Prenylation of DmCpx7A may tether it to membranes, increasing its local concentration at sites of neurotransmitter release and allowing it to bind to more SNARE complexes per vesicle. To function effectively as a fusion clamp, we hypothesize that Cpx would need to prevent even a single SNARE complex at the synaptic vesicle-membrane interface from exerting fusion-triggering forces. On the other hand, unprenylated DmCpx7B might be present at a lower local concentration at sites of neurotransmitter release. This could lead to binding of DmCpx7B to fewer SNARE complexes per vesicle. If the ability of Cpx to inhibit spontaneous neurotransmitter release were directly correlated to the number of SNARE complexes bound per vesicle, DmCpx7A would be expected to be a stronger fusion clamp than DmCpx7B, as we observe in our rescue experiments (Fig. 7). Several possibilities exist for the enhanced evoked release in DmCpx7B expressing animals. If evoked fusion requires Ca2+-activated Syt 1 to bind to SNARE complexes and displace Cpx, a reduced amount of DmCpx7B attached to SNARE complexes would make it easier for Syt 1 to trigger fusion. Moreover, the ability of Cpx to promote evoked neurotransmitter release has been linked to protein mobility. In vitro, GPI-anchored mCpx 1 was found to inhibit SNARE-mediated cell fusion to a much greater degree than soluble mCpx 1 (Giraudo et al., 2006). More recently, real-time observation of fluorescently labeled Cpx during single evoked exocytic events in PC12 cells showed that Cpx was present only briefly at sites of exocytosis at the onset of fusion. A truncated form of Cpx, however, appeared ~0.5 sec prior to the onset of fusion and lingered for ~2 sec after fusion, and this longer dwell time was correlated with reduced transmitter release (An et al., 2010). If prenylation of DmCpx7A increased its dwell time at active zone release sites compared to unprenylated DmCpx7B, this might explain why DmCpx7B was better able to promote evoked neurotransmitter release than DmCpx7A in our rescue experiments. In summary, although Drosophila contains only one Cpx gene, compared to four Cpx genes in mammals, the DmCpx locus can produce multiple Cpx isoforms through alternative splicing and RNA editing, perhaps to serve the distinct needs of different synapses. We also demonstrate that alternative splicing of the C-terminus of DmCpx results in two isoforms with differing effects on spontaneous and evoked neurotransmitter release, highlighting the importance of the C-terminus in regulating the effect of Cpx on these two modes of synaptic vesicle fusion.

Fig. 8
Model for the differential effects of DmCpx7A and DmCpx7B on neurotransmitter release. A top-down view of a synaptic vesicle fusing with the plasma membrane driven by five SNARE complexes (orange) is shown. Cpx molecules bound to the SNARE complexes are ...


Drosophila genetics and molecular cloning

Drosophila melanogaster were cultured on standard medium at 25°C. Genomic evaluation of Drosophila cpx regions made use of Flybase (Tweedie et al., 2009). DmCpx7A was cloned from a Drosophila cDNA clone obtained from the Drosophila Genomics Research Center (Clone ID GH27718). DmCpx7B was cloned from a first-strand cDNA synthesized using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) from total RNA purified from Canton S adults using the RNeasy Mini Kit (Qiagen). PCR products were subcloned into the pValum vector downstream of a 10xUAS cassette. UAS-DmCpx7A, UAS-DmCpx7B, UAS-GFP-DmCpx7A, UAS-mCherry-DmCpx7A, UAS-GFP-DmCpx7B and UAS-mCherry-DmCpx7B were injected by Genetic Services Inc. (Cambridge, MA) into y1v1;;P{CaryP}attP2 embryos for a third chromosome insertion as described (Ni et al., 2008). UAS lines were then recombined into the cpxSH1 null mutant background (Huntwork and Littleton, 2007). The elavC155-Gal4 driver was used for pan-neuronal expression of transgenes. The cpx572 mutant was generated by feeding EMS to Canton S males and testing non-complementation of 5000 mutagenized lines with cpxSH1.

Immunostaining and western blot analysis

Immunostaining was performed on wandering 3rd instar larvae after rearing at 25°C as described previously (Huntwork and Littleton, 2007; Rieckhof et al., 2003). Anti-DmCpx antiserum (1:500) (Huntwork and Littleton, 2007) with a secondary antibody conjugated to Alexa-488 (Molecular Probes, Carlsbad, CA) and goat anti-HRP antiserum conjugated to DyLight 549 (Jackson ImmunoResearch) (1:1000, Jackson ImmunoResearch, West Grove, PA) were used for immunostaining. Immunoreactive proteins were visualized on a Zeiss Pascal Confocal with PASCAL software (Carl Zeiss MicroImaging, Inc.).

Western blotting of whole adult head lysates was performed using standard laboratory procedures with anti-DmCpx (1:5000). For quantitative western analysis, adult heads were collected and lysed in 10 µl of 4× Laemmli sample buffer/head (8% SDS, 40% glycerol, 20% 2-mercaptoethanol, 0.02% bromophenol blue, 250 mM Tris HCl, pH 6.8). The equivalent of one head was loaded per lane. Equal loading was assayed using anti-arginine kinase at 1:5000. Western blots were imaged using an Odyssey infrared scanner (Li-Cor). The intensity of each DmCpx band was standardized to the intensity of the corresponding arginine kinase band using Odyssey software (Li-Cor). Samples were prepared and loaded in triplicate for a total of nine measurements.

FRAP analysis

Larvae expressing DmCpx7A-GFP or DmCpx7B-GFP were preselected for FRAP experiments based on GFP intensity. 3rd instar larvae were dissected in HL3 solution and pinned on silgard plates. Nerves were cut to eliminate muscle contractions. NMJs were imaged under 40× magnification using a Perkin Elmer spinning disk confocal microscope equipped with a Hamamatsu C9100-13 camera. Areas of the bouton were bleached using a 50mW 488 laser operating on maximum power. Confocal images were collected before and after bleaching and fluorescence recovery was followed for 1 minute with 10 sec intervals. Total fluorescence of the bleached areas was measured before and after bleaching. Fluorescence recovery was measured as delta F between bleached and the recovering time point and plotted as a percentage of the initial fluorescence.

In situ hybridization

Sense and antisense RNA probes labeled with digoxigenin (Roche) were generated by in vitro transcription of the first 500 bp in the 3’ untranslated regions of both Cpx7A and Cpx7B. In situ hybridization was carried out as described previously (Tautz and Pfeifle, 1989).

Quantitative RT-PCR

For absolute quantification of DmCpx7A and DmCpx7B levels, total RNA was extracted from ten adult Canton S flies or 3rd instar larvae using an RNeasy Mini Kit (Qiagen) and treated with DNase I (Ambion) according to the manufacturers’ instructions. DNase I was then removed from each sample with the RNeasy Mini Kit. Single stranded cDNA was synthesized in a total volume of 20 µL from 1 µg of total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to the manufacturer’s protocol. PCR was carried out in quadruplicate for each of three independent total RNA samples per genotype in optical 96-well plates (Applied Biosystems) for a total of 12 measurements per data point. The reaction mixtures were as follows: 25 µL of 2× QuantiTect SYBR Green PCR Master Mix (Qiagen), 300 nM forward primer, 300 nM reverse primer, and 5 µL of single stranded cDNA (see above) in a total volume of 50 µµL. The thermal cycling conditions were 15 min at 95°C, followed by 40 cycles of 15 sec at 94°C, 30 sec at 54°C, and 30 sec at 72°C. A final dissociation step was carried out to evaluate product integrity, and reaction samples were run on a 1.2% agarose gel and stained with ethidium bromide. The absolute levels of DmCpx7A and DmCpx7B were calculated using SDS software (Applied Biosystems) by comparison to a standard curve of DmCpx7A and DmCpx7B mRNA, respectively, transcribed in vitro. Values were normalized to the level of DmCpx7A transcripts in adult flies. For relative quantification of mRNA levels in the activity-dependence experiments (Fig. 2D), the levels of DmCpx7A and DmCpx7B were calculated using SDS software (Applied Biosystems) by the 2−ΔΔCt method using actin (Act88F) as an internal control and untreated wild-type (Canton S) animals as a calibrator. Samples were prepared in triplicate. The primer sequences were as follows: Act88F forward 5’-ACTTCTGCTGGAAGGTGGAC-3’ and reverse 5’-ATCCGCAAGGATCTGTATGC-3’, DmCpx7A forward 5’-CCCCCAAGAAGAGCCCAATC-3’ and reverse 5’-CACTGCATGACACATTTTCCCTCTAT-3’, and DmCpx7B forward 5’-CGCCGAAGCGGAGCAGGAAGAG-3’ and reverse 5’-GGGCGTGCTGGTGTGGGTGTCT-3’.

Electrophysiological analysis

Postsynaptic currents from the specified genotypes were recorded at segment A3 of ventral longitudinal muscle 6 in 3rd instar larvae using two-electrode voltage clamp with a -80 mV holding potential in modified HL3 solution (in mM: 10 NaHCO3, 5 KCl, 4 MgCl, 5 HEPES, 70 NaCl, 5 Trehalose, 115 Sucrose, pH 7.2). Final Ca2+ concentration was adjusted to the desired level indicated in the text. Data acquisition was performed using Axoscope 9.0 software. Quantal content was estimated by dividing the current integral of nerve-evoked currents by the current integral from individual quanta. Motor nerves innervating the musculature were severed and placed into a suction electrode so action potential stimulation could be applied at the indicated frequencies using a programmable stimulator. Electrophysiology analysis was performed using Clampfit 9.0 software (Axon Instruments, Foster City). Statistical analysis and graphs were performed using Origin Software (OriginLab Corporation, Northampton, MA, USA). Statistical significance was determined using a two-tailed Student’s t test. For all data, error bars represent s.e.m.


We thank Krys Foster and Amanda Hartman for help with EMS screening and Robin Stevens for supplying the Synaptogyrin-GFP transgenic line. RAJ was supported by the PEW Latin American Fellow Program in Biomedical Science. This work was supported by NIH grant NS40296 to J.T.L.


neuromuscular junction
excitatory junctional potential
soluble N-ethylmaleimide-sensitive fusion attachment protein receptor
Drosophila melanogaster
standard error of the mean
horse radish peroxidase
untranslated region


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • An SJ, Grabner CP, Zenisek D. Real-time visualization of complexin during single exocytic events. Nature Neuroscience. 2010;13:577–583. [PMC free article] [PubMed]
  • Bracher A, Kadlec J, Betz H, Weissenhorn W. X-ray structure of a neuronal complexin-SNARE complex from squid. J Biol Chem. 2002;277:26517–26523. [PubMed]
  • Brand AH, Perrimon N. Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development. 1993;118:401–415. [PubMed]
  • Cai H, Reim K, Varoqueaux F, Tapechum S, Hill K, Sørensen JB, Brose N, Chow RH. Complexin II plays a positive role in Ca2+-triggered exocytosis by facilitating vesicle priming. Proc Natl Acad Sci USA. 2008;105:19538–19543. [PubMed]
  • Chen X, Tomchick DR, Kovrigin EcDA, Machius M, Südhof TC, Rizo J. Three-dimensional structure of the complexin/SNARE complex. Neuron. 2002;33:397–409. [PubMed]
  • Cho RW, Song Y, Littleton JT. Comparative analysis of Drosophila and mammalian complexins as fusion clamps and facilitators of neurotransmitter release. Mol Cell Neurosci. 2010;45:389–397. [PMC free article] [PubMed]
  • Geppert M, Goda Y, Hammer RE, Li C, Rosahl TW, Stevens CF, Südhof TC. Synaptotagmin I: a major Ca2+ sensor for transmitter release at a central synapse. Cell. 1994;79:717–727. [PubMed]
  • Giraudo CG, Eng WS, Melia TJ, Rothman JE. A clamping mechanism involved in SNARE-dependent exocytosis. Science. 2006;313:676–680. [PubMed]
  • Giraudo CG, Garcia-Diaz A, Eng WS, Chen Y, Hendrickson WA, Melia TJ, Rothman JE. Alternative zippering as an on-off switch for SNARE-mediated fusion. Science. 2009;323:512–516. [PMC free article] [PubMed]
  • Giraudo CG, Garcia-Diaz A, Eng WS, Yamamoto A, Melia TJ, Rothman JE. Distinct domains of complexins bind SNARE complexes and clamp fusion in vitro. J Biol Chem. 2008;283:21211–21219. [PMC free article] [PubMed]
  • Guan Z, Saraswati S, Adolfsen B, Littleton JT. Genome-wide transcriptional changes associated with enhanced activity in the Drosophila nervous system. Neuron. 2005;48:91–107. [PubMed]
  • Higuchi M, Single FN, Köhler M, Sommer B, Sprengel R, Seeburg PH. RNA editing of AMPA receptor subunit GluR-B: a base-paired intron-exon structure determines position and efficiency. Cell. 1993;75:1361–1370. [PubMed]
  • Hobson RJ, Liu Q, Watanabe S, Jorgensen EM. Complexin Maintains Vesicles in the Primed State in C. elegans. Curr Biol. 2011;21:106–113. [PMC free article] [PubMed]
  • Hoopengardner B, Bhalla T, Staber C, Reenan R. Nervous system targets of RNA editing identified by comparative genomics. Science. 2003;301:832–836. [PubMed]
  • Huntwork S, Littleton JT. A complexin fusion clamp regulates spontaneous neurotransmitter release and synaptic growth. Nat Neurosci. 2007;10:1235–1237. [PubMed]
  • Kaeser-Woo YJ, Yang X, Sudhof TC. C-terminal complexin sequence is selectively required for clamping and priming but not for Ca2+ triggering of synaptic exocytosis. J. Neurosci. 2012;32:2877–2885. [PMC free article] [PubMed]
  • Karatekin E, Giovanni JD, Iborra C, Coleman J, O’Shaughnessy B, Seagar M, Rothman JE. A fast, single-vesicle fusion assay mimics physiological SNARE requirements. Proc Natl Acad Sci USA. 2010;107:3517–3521. [PubMed]
  • Littleton JT, Bellen HJ, Perin MS. Expression of synaptotagmin in Drosophila reveals transport and localization of synaptic vesicles to the synapse. Development. 1993;118:1077–1088. [PubMed]
  • Malsam J, Seiler F, Schollmeier Y, Rusu P, Krause J, Söllner T. The carboxy-terminal domain of complexin I stimulates liposome fusion. Proc Natl Acad Sci USA. 2009;106:2001–2006. [PubMed]
  • Martin JA, Hu Z, Fenz KM, Fernandez J, Dittman JS. Complexin has opposite effects on two modes of synaptic vesicle fusion. Curr Biol. 2011;21:97–105. [PMC free article] [PubMed]
  • Maximov A, Tang J, Yang X, Pang ZP, Südhof TC. Complexin controls the force transfer from SNARE complexes to membranes in fusion. Science. 2009;323:516–521. [PMC free article] [PubMed]
  • McMahon HT, Missler M, Li C, Südhof TC. Complexins: cytosolic proteins that regulate SNAP receptor function. Cell. 1995;83:111–119. [PubMed]
  • Mohrmann R, Wit H, de Verhage M, Neher E, Sørensen JB. Fast vesicle fusion in living cells requires at least three SNARE complexes. Science. 2010;330:502–505. [PubMed]
  • Pabst S, Hazzard JW, Antonin W, Südhof TC, Jahn R, Rizo J, Fasshauer D. Selective interaction of complexin with the neuronal SNARE complex. Determination of the binding regions. J Biol Chem. 2000;275:19808–19818. [PubMed]
  • Pickering BM, Willis AE. The implications of structured 5' untranslated regions on translation and disease. Semin Cell Dev Biol. 2005;16:39–47. [PubMed]
  • Reim K, Mansour M, Varoqueaux F, McMahon HT, Südhof TC, Brose N, Rosenmund C. Complexins regulate a late step in Ca2+-dependent neurotransmitter release. Cell. 2001;104:71–81. [PubMed]
  • Reim K, Wegmeyer H, Brandstätter JH, Xue M, Rosenmund C, Dresbach T, Hofmann K, Brose N. Structurally and functionally unique complexins at retinal ribbon synapses. J Cell Biol. 2005;169:669–680. [PMC free article] [PubMed]
  • Rieckhof GE, Yoshihara M, Guan Z, Littleton JT. Presynaptic N-type calcium channels regulate synaptic growth. J Biol Chem. 2003;278:41099–41108. [PubMed]
  • Schaub JR, Lu X, Doneske B, Shin Y-K, McNew JA. Hemifusion arrest by complexin is relieved by Ca2+-synaptotagmin I. Nat Struct Mol Biol. 2006;13:748–750. [PubMed]
  • Seiler F, Malsam J, Krause JM, Söllner TH. A role of complexin-lipid interactions in membrane fusion. FEBS Lett. 2009;583:2343–2348. [PMC free article] [PubMed]
  • Strenzke N, Chanda S, Kopp-Scheinpflug C, Khimich D, Reim K, Bulankina AV, Neef A, Wolf F, Brose N, Xu-Friedman MA, et al. Complexin-I is required for high-fidelity transmission at the endbulb of Held auditory synapse. J. Neurosci. 2009;29:7991–8004. [PubMed]
  • Südhof TC, Rothman JE. Membrane fusion: grappling with SNARE and SM proteins. Science. 2009;323:474–477. [PMC free article] [PubMed]
  • Tautz D, Pfeifle C. A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback. Chromosoma. 1989;98:81–85. [PubMed]
  • Tweedie S, Ashburner M, Falls K, Leyland P, McQuilton P, Marygold S, Millburn G, Osumi-Sutherland D, Schroeder A, Seal R, Zhang H. FlyBase Consortium. Flybase: enhancing Drosophila gene ontology annotations. Nucleic Acids Res. 2009;37:D555–D559. [PMC free article] [PubMed]
  • van den Bogaart J, Holt MG, Bunt G, Riedel D, Wouters FS, Jahn R. One SNARE complex is sufficient for membrane fusion. Nat. Struct. Mol. Biol. 2010;17:358–364. [PMC free article] [PubMed]
  • Xu J, Mashimo T, Südhof TC. Synaptotagmin-1, -2, and -9: Ca2+ Sensors for Fast Release that Specify Distinct Presynaptic Properties in Subsets of Neurons. Neuron. 2007;54:567–581. [PubMed]
  • Xue M, Craig TK, Xu J, Chao H-T, Rizo J, Rosenmund C. Binding of the complexin N terminus to the SNARE complex potentiates synaptic-vesicle fusogenicity. Nature Structural & Molecular Biology. 2010;17:568–575. [PMC free article] [PubMed]
  • Xue M, Lin YQ, Pan H, Reim K, Deng H, Bellen HJ, Rosenmund C. Tilting the balance between facilitatory and inhibitory functions of mammalian and Drosophila Complexins orchestrates synaptic vesicle exocytosis. Neuron. 2009;64:367–380. [PMC free article] [PubMed]
  • Xue M, Reim K, Chen X, Chao H-T, Deng H, Rizo J, Brose N, Rosenmund C. Distinct domains of complexin I differentially regulate neurotransmitter release. Nat Struct Mol Biol. 2007;14:949–958. [PMC free article] [PubMed]
  • Xue M, Stradomska A, Chen H, Brose N, Zhang W, Rosenmund C, Reim K. Complexins facilitate neurotransmitter release at excitatory and inhibitory synapses in mammalian central nervous system. Proc Natl Acad Sci USA. 2008;105:7875–7880. [PubMed]
  • Yoshihara M, Littleton JT. Synaptotagmin I Functions as a Calcium Sensor to Synchronize Neurotransmitter Release. Neuron. 2002;36:897–908. [PubMed]
  • Zhang FL, Casey PJ. Protein prenylation: molecular mechanisms and functional consequences. Annu Rev Biochem. 1996;65:241–269. [PubMed]
  • Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 2003;31:3406–3415. [PMC free article] [PubMed]