Neuronal communication is made possible by the release of neurotransmitters, which in turn depends on the fusion of neurotransmitter-laden synaptic vesicles at the ends of nerve cells. Synaptic vesicle fusion is triggered by an influx of Ca2+ ions into the neuron upon depolarization of the neuron, a process that initiates neurotransmission. Neurotransmitter release is quantized; that is, at most one synaptic vesicle fuses in the active zone upon an action potential. This process is controlled by several proteins, including SNAREs (soluble NSF [N-ethylmaleimide-sensitive factor] attachment protein receptors), the Ca2+sensor synaptotagmin 1, Munc18, complexin, and the ATPase NSF, among others. Thus, neurotransmitter release is a biological phenomenon controlled by complex interactions between individual molecules. An understanding of the underlying molecular mechanisms requires methods that are capable of observing single vesicles and molecules.
Ideally, observations of single vesicles and molecules would be performed in live neurons. Although such studies have been under way (in our laboratory and others), they currently provide limited information, largely because the genetic manipulations and labeling techniques used do not provide the spatial and time resolution required for studying the molecular mechanism of neurotransmitter release. Thus, minimal in vitro systems are needed that mimic the neurotransmitter release characteristics observed in neurons and that allow manipulations and observations not possible in vivo. Such in vitro systems will set the stage for deciphering of the effect of other factors on the process. They could also become screening tools for the development of therapeutic leads to modulate neurotransmitter release and combat neurological disorders.
Our approach to understanding the molecular basis for neurotransmitter release consists of a combination of structural and biophysical studies of in vitro systems. Structural information about complexes between the individual molecular components is primarily obtained by x-ray crystallography and single-molecule FRET (fluorescence resonance energy transfer) measurements; information about membrane morphology is obtained by cryo-electron microscopy. This hybrid approach provides the framework for investigations targeted at the functional and dynamic aspects of the system, using single-molecule and single-particle fluorescence microscopy techniques.
In Vitro Reconstitution of Ca2+-Triggered Synaptic Vesicle Fusion
We have achieved the long-standing goal to establish an in vitro single vesicle-vesicle system with reconstituted synaptic proteins that produces a rapid burst of content release upon injection of Ca2+. Donor vesicles mimic synaptic vesicles, and acceptor vesicles mimic the plasma membrane in the active zone. Prior to Ca2+ injection, the system is in a state of single interacting pairs of donor and acceptor vesicles, and fusion events are rare. Our system differentiates between membrane docking, membrane lipid exchange, and complete fusion (i.e., pore formation) upon Ca2+ injection, the latter mimicking quantized neurotransmitter release upon exocytosis of synaptic vesicles. Initially, we included neuronal SNARE, synaptotagmin 1, and complexin, and we are currently investigating the function of other proteins with our system, including Munc18 and NSF/SNAP. We have also begun to use our system to investigate the effect of α-synuclein, a factor that is implicated in Parkinson's disease.
With our single vesicle-vesicle system we discovered the temporal sequence of membrane states for fusion on a 100-millisecond timescale upon Ca2+ injection (in the 250- to 500-μM range) with reconstituted SNARE and synaptotagmin-1 proteoliposomes. In addition, we imaged detailed membrane morphologies with cryo-electron microscopy before and after Ca2+ injection. We observed a heterogeneous network of fast and slow fusion pathways. Remarkably, all instances of Ca2+-triggered fast fusion started from a membrane-membrane point contact and proceeded to complete fusion within less than 100 milliseconds, without discernible intermediates. In contrast, pathways that involved a stable hemifusion diaphragm only resulted in fusion after seconds, if any. When complexin was included in our system, the network of Ca2+-triggered fusion pathways was shifted toward the fast pathway, effectively synchronizing fusion, especially at lower Ca2+ concentration. Synaptic proteins may have evolved to select this fast, immediate pathway out of a heterogeneous network of possible membrane fusion pathways.
On a more technical note, our system is a major advance over previous in vitro assays that have been commonly used in the SNARE field. Because our system includes both lipid- and content-mixing reporting fluorophores, and because we use total internal reflection light microscopy to monitor single vesicles, different events have characteristic fluorescent signals and can be distinguished on this basis. We can discriminate between docking, hemifusion (characterized by just lipid mixing), and complete fusion (i.e., fusion pore opening) for each individual vesicle. In contrast to ensemble measurements in which single vesicles are not monitored—and which is how such reconstituted systems have been typically monitored in the past—this also allows us to distinguish between fusing vesicles and those that burst or leak accidentally. Moreover, most previous experiments in the field of SNARE-mediated fusion only monitored lipid mixing that is necessary but not sufficient for complete fusion. Our results suggest that some of the conclusions of these earlier studies need to be revised by monitoring content mixing.
A Structural Model of the Prefusion State
Based on the available data from our laboratory and others, we proposed a model (Figure 1) for Ca2+-triggered vesicle fusion that begins with a SNARE/synaptotagmin-induced point contact between membranes, without initial hemifusion. The Ca2+-binding loops of synaptotagmin are close to, but yet not penetrating, the membrane. Upon Ca2+ influx, their Ca2+-binding loops insert into the membrane. Simultaneously, the synaptotagmin-SNARE interaction may lead to full zippering of the SNARE complex or exert a force on the transmembrane domains. In concert, the interaction of the synaptotagmin Ca2+-binding loops with the membrane perturbs membrane regions near the point contact. Together these processes set the stage for fast fusion in the millisecond timescale. We are currently investigating how complexin might interact with this synaptotagmin-SNARE prefusion complex.
Effect of α-Synuclein, a Protein Implicated in Parkinson's Disease
Mutations in the gene that codes for α-synuclein, or gene duplications and triplications, are known to lead to an increased risk of early-onset Parkinson's disease, a condition where the central nervous system degenerates and where α-synuclein aggregates form clumps called Lewy bodies in the neurons of patients. α-Synuclein is also thought to have a role in neurotransmitter release, but the details of its involvement are not fully understood.
We used our single-vesicle technology to study these questions. In collaboration with Thomas Südhof (HHMI, Stanford University), we found that native α-synuclein (in conjunction with SNAREs) increases the availability of vesicles at the synapse by causing them to cluster together. In contrast, and in agreement with in vivo data, native α-synuclein did not interfere with Ca2+-triggered fusion in our single-vesicle system. Thus, α-synuclein increases the local concentration of synaptic vesicles at the synapse but does not affect Ca2+-triggered fusion per se. In contrast, it is known that pathogenic α-synuclein aggregates directly interfere with the release of the neurotransmitter molecules. In fact, large α-synuclein oligomers interfere with SNARE complex formation. Moreover, when we used a particular mutant form of α-synuclein that is associated with Parkinson's disease (A30P), the synaptic vesicle mimics did not form clusters. If these results are confirmed in vivo, the role played by native α-synuclein in the central nervous system and the connection between α-synuclein and Parkinson's disease will be clearer. We are planning mechanistic structure/function studies using single-molecule FRET to map the molecular interactions between synaptobrevin and α-synuclein. We will then use rescue experiments in the triple-knockout mice created by the Südhof laboratory to probe these interactions.
Synapse formation and maturation are essential for the normal establishment and remodeling of neuronal circuits in the brain. Impairments in synapse formation and maturation are causes of human diseases such as autism spectrum disorders and mental retardation. Among trans-synaptic adhesion factors that play a role in synapse formation are the neurexin and neuroligin family of proteins. Neuroligins and neurexins form a complex in which two neuroligin molecules link to each other, and a neurexin molecule attaches to each side of the pair. Complex formation is downstream from cadherin and SynCAM adhesion protein engagement. As neurons create new synapses during learning, they must form neuroligin-neurexin connections for those synapses to become functionally mature.
To better understand how the two proteins interact, my laboratory determined the structure of the neuroligin-1 and neurexin-1β complex (Figure 2). We are currently investigating other complexes between neuroligins, neurexins, and additional ligands.
Advanced Biomolecular Imaging at the Molecular Scale
X-ray diffraction—which reveals atomic structures of proteins, nucleic acids, and their complexes—plays a pivotal role in the understanding of biological systems. Recently there has been much interest in very large assemblies, such as the ribosome. Since crystals of such large assemblies often diffract weakly (resolution worse than ~3.5 Å), there is a need to develop methods that work at such low resolution. In macromolecular assemblies, some of the components may be known at high resolution, while others are unknown. Determining the structure of such complexes, which are often biologically important, should be possible in principle, as the number of independent diffraction intensities at a resolution better than ~5 Å generally exceeds the number of degrees of freedom.
The availability of new powerful x-ray sources promises to open the door for structure determination of challenging biological systems that cannot be achieved with "conventional" synchrotrons, or, at the minimum, to extend the resolution of such structures. The world's first hard x-ray free-electron laser, the Linac Coherent Light Source (LCLS) at the SLAC National Accelerator Laboratory (SLAC), provides coherent x-rays in a pulse duration in the tens of femtoseconds with a peak brightness roughly 10 orders of magnitude higher than is available on the world's most powerful synchrotron x-ray sources. This facility may offer opportunities for structural studies of large biomolecular complexes and subcellular structures.
To aid with the analysis of low-resolution diffraction data from LCLS as well as conventional synchrotrons, we developed a method (deformable elastic network [DEN] refinement) that adds specific information from known homologous structures but allows global and local deformations of these homology models. Our approach uses the observation that local protein structure tends to be conserved as sequence and function evolve. For test cases at 3.5- to 7-Å resolution, with known structures at high resolution, our method is a significant improvement over conventional refinement, as monitored by coordinate accuracy, the definition of secondary structure, and the quality of electron density maps. Our method was instrumental in the determination of a number of new x-ray crystal structures (carried out by us and by other groups), and it is applicable to the study of weakly diffracting crystals with a high degree of anisotropy or thermal factors.
Further developments in this area are part of a recently funded Hughes Collaborative Innovation Award to exploit the potential of LCLS, enabling data collection from submicron- to micron-sized crystals without the deleterious influence of radiation damage, using the extremely large photon flux within the femtosecond x-ray pulse of LCLS to capture the Bragg diffraction pattern before the crystal is destroyed. We are planning the necessary methodological developments for sample delivery, data collection, phasing, and refinement with LCLS diffraction data and their application to several problems in the forefront of biological structure determination, in collaboration with James Berger (University of California, Berkeley), David Eisenberg (HHMI, University of California, Los Angeles), Douglas Rees (HHMI, California Institute of Technology), and William Weis (Stanford University).
A grant from the National Institutes of Health provides partial support for the work on neurotransmitter release.
As of August 21, 2013