EuroSYNAPSE - Spatio-temporal organization of the synaptic membrane for synaptic vesicle protein recycling

突触前终末含有突触囊泡(SVS)簇，这是化学神经传递的关键细胞器(1)。蛋白质组学数据表明，SVS由不同的蛋白质和脂类组成，这些蛋白质和脂类存在于特定的化学计量学(2)中，在重复的外吞和内吞过程中必须保持这些蛋白质和脂类。这种对SV膜蛋白的精确分选是如何从分子上完成的还不是很清楚。如高分辨率受激发射耗竭(STED)显微镜(3)所示，SV组分在胞外过程中可以保持聚集，或者通过货物特异性接头蛋白(如AP-2、Stonin2/Stoned B(5，6)、AP180等)对分选决定因素的识别而在细胞表面(4)重新聚集。SV外吞和内吞在时间上是耦合的，但在活动区(融合部位)和周围的内吞或周围活动区之间是空间隔离的。SV外吞和内吞之间的紧密时间耦合(7)表明，SV货物蛋白的分类和回收可能涉及精确控制内吞蛋白或蛋白质复合体在融合部位(活动区)和内吞(周边区)之间的定位和动态(8)。神经末梢的内吞蛋白的时空动态如何被控制在很大程度上是未知的，但可能涉及组织周围活动区的膜相关多结构域支架蛋白(7，9)。虽然近年来在识别和表征聚集在活动区(CAZ，也称为突触前网格)的Cytomatrix方面取得了巨大的进展，但对其周围活动区的结构成分和功能组织知之甚少(7)。从简化的角度来看，周围活动区可以被视为一个三层结构，由表面搁置的SV货物池、内吞蛋白的动态组装和基本固定的支架组成，例如巨大的蛋白质短笛。然而，该组织的确切基本机制尚未解开。在SV循环过程中，内吞蛋白进出周围活动区的运动伴随着网状蛋白涂层结构的伴随组装。这一过程涉及蛋白质-蛋白质和蛋白质-脂质相互作用的调节网络的协调形成。内吞网络似乎是围绕着中央枢纽组织的，即在给定的时间或限定的空间内显示出不成比例的高数量相互作用的因素(10)。随着内吞囊泡的成熟，网络从SV货物和磷脂酰肌醇(4，5)-二磷酸(PIP2)-进入AP-2-，最后进入以笼蛋白为中心的状态(6，10)。许多研究表明，这一过程涉及到内吞蛋白或其复合体被调控地重新定位到SV内吞作用的部位(1，8)。超微结构和光学显微镜数据结合显性-负性方法表明，分子上类似的成分，如SH3结构域包含的蛋白质两体素、突触素、内亲素和交叉素被招募到周围活动区，并在SV循环中发挥不重叠的功能作用。基于遗传、生化和功能的综合研究，胞内蛋白质可分为功能蛋白质模块，即由它们在特定的时间域内的物理相互作用和它们的空间隔离定义的大分子复合体(11)。例如，支架蛋白交叉素在物理上和功能上与Eps15、Dynamin和Synaptojanin相互作用。在黑腹毛虫中，Eps15和Intersectin的突变体表现出彼此相似的表型(12)，这两种蛋白都是根据活性从突触周围中心共同迁移到周围活动区的，这表明Intersectin与Eps15一起是一个功能模块的一部分，该功能模块在clathrin介导的SV内吞作用的后期阶段调节Dynamin和Synaptojanin的定位和活性(图1B)。目前，我们还不清楚这种运动是如何在分子水平上完成的，但可能涉及到与肌动蛋白细胞骨架的相互作用。由于内吞蛋白可以通过它们的EH结构域与几个模块联系在一起，因此它们的EH结构域也是一个功能模块的一部分，该模块包含突触凝集素特异的分选适配器StonedB/Stonin2、其相互作用元件GIT1(我们自己的未发表数据)、AP180和AP-2复合体，这些都与从突触前质膜中检索表面链的SV货物池有关(图1A)。因此，功能性内吞蛋白模块可以被视为处于严格时空控制下的周围活动区的动态组件。这种调控是如何完成的，以及它如何在融合后与囊泡膜和相关组件的迁移同步仍是未知的(9)，这主要是因为目前的方法允许在不进入时间域的情况下进行亚细胞成像(即通过传统的电子显微镜)，或者基于不能提供足够空间分辨率的光学显微镜技术的快速活细胞成像。这项建议中描述的工作旨在解决这些问题，从而使我们能够获得前所未有的洞察突触膜的时空组织，以实现网状蛋白介导的SV蛋白循环。

The presynaptic terminal contains clusters of synaptic vesicles (SVs), key organelles of chemical neurotransmission (1). Proteomics data indicate that SVs comprise distinct sets of proteins and lipids present in defined stoichiometries (2) which must be maintained during repetitive rounds of exo- and endocytosis. How this precise sorting of SV membrane proteins is accomplished molecularly is not well understood. SV components could remain clustered during their exo-endocytic itinerary as suggested by high-resolution stimulated emission-depletion (STED) microscopy (3), or be reclustered at the cell surface (4) by individual sorting via recognition of sorting determinants by cargo-specific adaptor proteins such as AP-2, stonin2/stoned B (5, 6), AP180 and perhaps others. SV exo- and endocytosis appear to be temporally coupled but spatially segregated between the active zone (sites of fusion) and the surrounding endocytic or periactive zone. The tight temporal coupling between SV exo- and endocytosis (7) suggests that SV cargo protein sorting and recycling likely involves precise control of the localization and dynamics of endocytic proteins or protein complexes as they partition between sites of fusion (active zone) and endocytosis (periactive zone) (8). How the spatio-temporal dynamics of endocytic proteins at the nerve terminal are controlled is largely unknown but likely involves membrane-associated multidomain scaffolding proteins organizing the periactive zone (7, 9). While recent years have witnessed enormous progress in the identification and characterization of the cytomatrix assembled at the active zone (CAZ, also termed presynaptic grid) comparably little is known about the structural components and the functional organization of the periactive zone that surrounds it (7). From a simplified perspective the periactive zone might be viewed as a three-layered structure comprised of a pool of surface stranded SV cargo, dynamic assemblies of endocytic proteins, and largely immobile scaffolds such as the giant protein piccolo. However, the precise underlying mechanisms for this organization have not been unravelled. Movement of endocytic proteins to and from the periactive zone during SV cycling is accompanied by the concomitant assembly of clathrin-coated structures. This process involves the coordinated formation of a regulated network of protein-protein and protein-lipid interactions. The endocytic network appears to be organized around central hubs, i.e. factors that at a given time or within a defined space display a disproportionately high number of interactions (10). As maturation of the endocytic vesicle progresses, the network proceeds from a SV cargo and phosphatidylinositol (4,5)-bisphosphate (PIP2)- to an AP-2-, and finally to a clathrin-centered state (6, 10). A number of studies indicate that this progression involves the regulated relocalization of endocytic proteins or complexes thereof to sites of SV endocytosis (1, 8). Ultrastructural and light microscopy data in combination with dominant-negative approaches have revealed that molecularly similar components such as the SH3 domain containing proteins amphiphysin, syndapin, endophilin, and intersectin are recruited to the periactive zone and play non-overlapping functional roles in SV recycling. Based on combined genetic, biochemical, and functional studies endocytic proteins might be grouped into functional protein modules, i.e. macromolecular complexes defined by their physical interactions within a characteristic time domain and by their spatial sequestration (11). For example, the scaffolding protein intersectin physically and functionally interacts with eps15, dynamin, and synaptojanin. Mutants of eps15 and intersectin in D. melanogaster phenotypically resemble each other (12) and both proteins co-migrate from the centre of synaptic boutons to the periactive zone in response to activity, suggesting that intersectin together with eps15 is part of a functional module that regulates localization and activity of dynamin and synaptojanin at late stages of clathrin-mediated SV endocytosis (Figure 1B). At present we do not understand how such movement is accomplished at the molecular level but interactions with the actin cytoskeleton are likely to be involved. As endocytic proteins can be associated with several modules intersectin and eps15 via their EH domains are also part of a functional module containing the synaptotagmin-specific sorting adaptor stonedB/ stonin2, its interactor GIT1 (our own unpublished data), AP180, and the AP-2 complex implicated in retrieval of surface-stranded pools of SV cargo from the presynaptic plasmalemma (Figure 1A). Thus, functional endocytic protein modules may be viewed as dynamic components of the periactive zone that are under tight spatio-temporal control. How such regulatory control is accomplished and how it is synchronized with the migration of the vesicle membrane and associated components after fusion remains unknown (9), largely because current methodology allows either subcellular imaging without access to the time domain (i.e. by conventional electron microscopy) or fast live cell imaging based on light microscopic techniques that do not provide sufficient spatial resolution. The work described in this proposal aims to tackle these problems and will thus enable us to gain unprecedented insights into the spatio-temporal organization of the synaptic membrane for clathrin-mediated SV protein recycling.