Membrane traffic in eukaryotic cells involves transport of vesicles that bud from a donor compartment and fuse with an acceptor compartment. Common principles of budding and fusion have emerged, and many of the proteins involved in these events are now known. However, a detailed picture of an entire trafficking organelle is not yet available. Using synaptic vesicles as a model, we have now determined the protein and lipid composition; measured vesicle size, density, and mass; calculated the average protein and lipid mass per vesicle; and determined the copy number of more than a dozen major constituents. A model has been constructed that integrates all quantitative data and includes structural models of abundant proteins. Synaptic vesicles are dominated by proteins, possess a surprising diversity of trafficking proteins, and, with the exception of the V-ATPase that is present in only one to two copies, contain numerous copies of proteins essential for membrane traffic and neurotransmitter uptake.

The TMEM16 family of proteins, also known as anoctamins, features a remarkable functional diversity. This family contains the long sought-after Ca(2+)-activated chloride channels as well as lipid scramblases and cation channels. Here we present the crystal structure of a TMEM16 family member from the fungus Nectria haematococca that operates as a Ca(2+)-activated lipid scramblase. Each subunit of the homodimeric protein contains ten transmembrane helices and a hydrophilic membrane-traversing cavity that is exposed to the lipid bilayer as a potential site of catalysis. This cavity harbours a conserved Ca(2+)-binding site located within the hydrophobic core of the membrane. Mutations of residues involved in Ca(2+) coordination affect both lipid scrambling in N. haematococca TMEM16 and ion conduction in the Cl(-) channel TMEM16A. The structure reveals the general architecture of the family and its mode of Ca(2+) activation. It also provides insight into potential scrambling mechanisms and serves as a framework to unravel the conduction of ions in certain TMEM16 proteins.

Uptake of glutamate into synaptic vesicles is mediated by vesicular glutamate transporters (VGLUTs). Although glutamate uptake has been shown to depend critically on Cl(-), the precise contribution of this ion to the transport process is unclear. We found that VGLUT1, and not ClC-3 as proposed previously, represents the major Cl(-) permeation pathway in synaptic vesicles. Using reconstituted VGLUT1, we found that the biphasic dependence of glutamate transport on extravesicular Cl(-) is a result of the permeation of this anion through VGLUT1 itself. Moreover, we observed that high luminal Cl(-) concentrations markedly enhanced loading of glutamate by facilitation of membrane potential-driven uptake and discovered a hitherto unrecognized transport mode of VGLUT1. Because a steep Cl(-) gradient across the synaptic vesicle membrane exists in endocytosed synaptic vesicles, our results imply that the transport velocity and the final glutamate content are highly influenced, if not determined, by the extracellular Cl(-) concentration.

observation that synaptic strength correlates with dendritic spine morphology leads to the hypothesis that the mushroom-like shape of dendritic spines functions as a receptor trap. We developed a mimetic system to investigate dendritic spine morphology and its effects on receptor confinement and diffusion. Giant unilamellar vesicles (GUV's) are made from lipids using electroswelling. To mimic the mushroom-shaped morphologies of dendritic spines, a micromanipulator is used to pull membrane tubes from the GUV lipid bilayer. Trapping capabilities for different spine morphologies are assessed by tracking quantum dots attached to membrane lipids, thus mimicking receptors. Results show a strong dependence of escape times on GUV morphology, as quantified by GUV radius and tube length. Instead of a trivial quadratic dependence of escape times on GUV radius we find a powerlaw dependence with an exponent of 2.85. This confirms the idea that receptors can be trapped by the morphology of a dendritic spine. Therefore the connection strength of a mushroom-shaped dendritic spine is much more stable than the strengths of stubby shaped dendritic spines.

The calcium-activated chloride channel TMEM16A is a member of a conserved protein family that comprises ion channels and lipid scramblases. Although the structure of the scramblase nhTMEM16 has defined the architecture of the family, it was unknown how a channel has adapted to cope with its distinct functional properties. Here we have addressed this question by the structure determination of mouse TMEM16A by cryo-electron microscopy and a complementary functional characterization. The protein shows a similar organization to nhTMEM16, except for changes at the site of catalysis. There, the conformation of transmembrane helices constituting a membrane-spanning furrow that provides a path for lipids in scramblases has changed to form an enclosed aqueous pore that is largely shielded from the membrane. Our study thus reveals the structural basis of anion conduction in a TMEM16 channel and it defines the foundation for the diverse functional behavior in the TMEM16 family.DOI: http://dx.doi.org/10.7554/eLife.26232.001