Fusion Pore Formation Observed during SNARE-Mediated Vesicle Fusion with Pore-Spanning Membranes

Abstract

Planar pore-spanning membranes (PSMs) have been shown to be a versatile tool to resolve docking and elementary steps of the fusion process with single large unilamellar vesicles (LUVs). However, in previous studies, we monitored only lipid mixing and did not gather information about the formation of fusion pores. To address this important step of the fusion process, we entrapped sulforhodamine B at self-quenching concentrations into LUVs containing the v-SNARE synaptobrevin 2, which were docked and fused with lipid-labeled PSMs containing the t-SNARE acceptor complex ΔN49 prepared on porous silicon substrates. By dual color spinning disc fluorescence microcopy with a time resolution of 20 ms, we could unambiguously distinguish between bursting vesicles and fusion pore formation. Owing to the aqueous compartment underneath the PSMs, vesicle bursting turned out to be an extremely rare event (< 0.01 %). From the time-resolved dual color fluorescence time traces, we were able to identify different fusion pathways including remaining three-dimensional postfusion structures with released content and flickering fusion pores. Our results on fusion pore formation and lipid diffusion from the PSM into the fusing vesicle let us conclude that the content release, i.e., fusion pore formation follows the merger of the two lipid membranes by only about 40 ms.

In vitro single vesicle fusion assays based on pore-spanning membranes: merits and drawbacks

“…From the mean square displacements, a mean diffusion coefficient of 0.42 ± 0.15 μm 2 /s was calculated. (Mühlenbrock et al 2020) Compared to the diffusion constant of syntaxin 1A in the f-PSM, the diffusion coefficient of the docked vesicles is by a factor of five smaller. A close contact between vesicle and membrane and the displacement of the intermembrane water layer as well as multiple interacting SNARE-complexes might be responsible for the reduced diffusion coefficient compared to that of a single SNARE.…”

“…In the centre of the pore, the PSMs are flat presumably as they experience a certain pre-stress in the mN/m-range (Janshoff and Steinem 2015 ; Kocun et al 2011 ; Kuhlmann et al 2014 ). It is well conceivable that this membrane geometry, as well as the broad distribution of the ΔN49 complex concentration in the individual GUVs (Mühlenbrock et al 2020 ) influence the observed differences in the protein distribution.…”

“…The SNARE-binding site of the ΔN49 complex was incubated with the soluble synaptobrevin 2 fragment (aa 1–96) that is known to block fusion before protein reconstitution. No docking of synaptobrevin 2-doped vesicles on these PSM was observed (Hubrich et al 2019 ; Mühlenbrock et al 2020 ).…”

“…Besides the preferential location of the proteins at the pore edges, also the overall mobility of the membrane components adds to the observed immobility of the docked vesicles on the s-PSMs. The diffusion coefficients of the lipids and peptides are by a factor of 2–4 lower on the s-PSMs compared to the f-PSMs (Kuhlmann et al 2017 ; Mühlenbrock et al 2020 ; Schwenen et al 2015 ). Moreover, we hypothesise that a conformal contact of the vesicle with the membrane on the gold-covered support with a large Hamacker constant, further immobilises the vesicles (Kuhlmann et al 2017 ).…”

“…The docked vesicle (I) fuses and transfers its content (II) across the PSM and into the second aqueous compartment, thus leading to an increase in fluorescence intensity visible in ROI 2. c Schematic illustration of the process shown in A and B. Adapted from(Mühlenbrock et al 2020) Different fusion pathways extracted from single vesicle fusion events with idealised fluorescence intensity time traces. Taken from…”