Complexin inhibits spontaneous release and synchronizes Ca2+-triggered synaptic vesicle fusion by distinct mechanisms

Abstract

Previously we showed that fast Ca(2+)-triggered vesicle fusion with reconstituted neuronal SNAREs and synaptotagmin-1 begins from an initial hemifusion-free membrane point contact, rather than a hemifusion diaphragm, using a single vesicle-vesicle lipid/content mixing assay (Diao et al., 2012). When complexin-1 was included, a more pronounced Ca(2+)-triggered fusion burst was observed, effectively synchronizing the process. Here we show that complexin-1 also reduces spontaneous fusion in the same assay. Moreover, distinct effects of several complexin-1 truncation mutants on spontaneous and Ca(2+)-triggered fusion closely mimic those observed in neuronal cultures. The very N-terminal domain is essential for synchronization of Ca(2+)-triggered fusion, but not for suppression of spontaneous fusion, whereas the opposite is true for the C-terminal domain. By systematically varying the complexin-1 concentration, we observed differences in titration behavior for spontaneous and Ca(2+)-triggered fusion. Taken together, complexin-1 utilizes distinct mechanisms for synchronization of Ca(2+)-triggered fusion and inhibition of spontaneous fusion.

Keywords: SNARE; complexin; membrane fusion; neurotransmitter release; synaptic vesicle fusion; synaptotagmin.

Figure 1.

Reconstituted single vesicle-vescile fusion assay. (A) Protein composition and labeling of proteoliposomes. (B) SDS-PAGE analysis of reconstituted vesicles mimicking the plasma membrane (PM) and synaptic vesicles (SV). (C) Schema of the extended single vesicle–vesicle content mixing assay. The shaded backgrounds indicate the three subsequent 1 min time periods of the procedure and the short intervals (5 s) for buffer exchanges. Representative content fluorescence intensity time traces are shown for an associated pair of SV and PM vesicles without any fusion (top trace), for an associated pair that undergoes spontaneous fusion, and for an associated pair that undergoes fusion after Ca²⁺-injection. The association of an SV vesicle to a surface-immobilized PM vesicle is characterized by fluorescence intensity increases during the 1 min period after initial SV vesicle loading. Spontaneous and Ca²⁺-triggered fusion events are characterized by a subsequent stepwise increase of the content dye fluorescence intensity during the respective 1 min observation periods. The time point of the arrival of Ca²⁺ is determined by the appearance of fluorescence intensity from soluble Cy5 dyes that are part of the injected solution (bottom trace). DOI: http://dx.doi.org/10.7554/eLife.03756.004

Figure 2.. Complexin-1 synchronizes Ca²⁺-triggered fusion and inhibits spontaneous fusion in concert with synaptotagmin-1 and neuronal SNAREs.

(A and B) Probability of fusion vs time upon 500 μM Ca²⁺-injection and spontaneous fusion, both in the presence of full-length synaptotagmin-1 and neuronal SNAREs, but with and without 2 μM complexin-1 (Cpx) as indicated. (A) The histograms of Ca²⁺-triggered fusion events (1 s time bin) were normalized by the number of associated SV vesicles and fit to an exponential decay function. The decay rate with and without complexin-1 is 0.43 s⁻¹ and 0.11 s⁻¹ respectively. (B) The histograms of spontaneous fusion events were normalized by the number of associated SV vesicles. (C) Domain structure of complexin-1. To probe the roles of the four domains of complexin-1 we used the truncation mutant Cpx_26–134 for the N-terminal domain, the truncation mutant Cpx_41–134 for the accessory α-helical domain, the Cpx_4M mutant for the SNARE binding domain, and the truncation mutant Cpx_1–88 for the C-terminal domain. We chose these particular truncations/mutations based on studies of cortical neuronal cultures (Maximov et al., 2009; Kaeser-Woo et al., 2012). (D–G) The bar graphs show the effects of complexin-1 and its mutants on SV-PM vesicle association (D), the number of spontaneous fusion events over the 1-min observation period divided by the number of associated SV vesicles (E), the decay rate of the histogram upon Ca²⁺-injection (F), and the amplitude of the first 1-sec time bin upon Ca²⁺-injection (G). Each value in panel G was normalized by the respective number of fusion events after Ca²⁺-injection. Moreover, only those single vesicle–vesicle fluorescence intensity traces that exhibited a distinct event during the association period were analyzed for fusion events during the subsequent observation periods. Error bars in D, E, G are standard deviations for 6–13 independent repeat experiments. Error bars in F are error estimates computed from the covariance matrix upon fitting the corresponding histograms with a single exponential decay function using a Levenberg-Marquardt technique. DOI: http://dx.doi.org/10.7554/eLife.03756.002

Figure 3.. Suppression of spontaneous fusion and synchronization of Ca²⁺-triggered fusion vs complexin-1 concentration (in concert with synaptotagmin-1 and neuronal SNAREs).

The bar graphs show the effect of wildtype complexin-1 at specified concentrations on SV-PM vesicle association (A), the number of spontaneous fusion events divided by the number of associated SV vesicles (B), the decay rate of the histogram of fusion events upon Ca²⁺-injection (C), and the amplitude of the first 1-sec time bin upon Ca²⁺-injection (D). Each value in panel D was normalized by the respective number of fusion events after Ca²⁺-injection. Moreover, only those single vesicle–vesicle fluorescence intensity traces that exhibited a distinct event during the association period were analyzed for fusion events during the subsequent observation periods. Error bars in A, B, D show the standard deviation for 6–10 independent repeat experiments. Error bars in C are error estimates computed from the covariance matrix upon fitting the corresponding histograms with a single exponential decay function using a Levenberg-Marquardt technique. ** indicates p<0.01 using the Student's t test. DOI: http://dx.doi.org/10.7554/eLife.03756.005

Figure 4.. Suppression of spontaneous fusion by complexin-1 with or without synaptotagmin-1.

The bar graphs summarize the effect of 2 μM complexin-1 on SV-PM vesicle association (A) and the number of spontaneous fusion events divided by the number of associated SV vesicles (B); the conditions that did not include synaptotagmin-1 used SV vesicles that were reconstituted with synaptobrevin-2 only. Moreover, only those single vesicle–vesicle fluorescence intensity traces that exhibited a distinct event during the association period were analyzed for spontaneous fusion events. DOI: http://dx.doi.org/10.7554/eLife.03756.007