Dynamic Structural Properties of Synapses

通过预包埋纳米金标记培养的分离海马神经元来研究活动期间PSD组分的运动。用高K+去极化通常用于去极化和激活神经元培养物，但也使用更接近生理刺激的其他方案，例如在不存在Mg 2+的情况下应用NMDA和用甘氨酸进行突触激活。现在，几乎所有的PSD的主要成分的抗体已经获得和测试的兼容性免疫电子显微镜，活动诱导的重组的PSD在分子水平上的一般图片开始出现。支架蛋白如PSD-95和GKAP在短期活性期间似乎相对稳定，并且构成位于突触后膜旁边的30 nm厚的层，其中受体通过与支架的结合而稳定。

相比之下，其他蛋白质在短期活动期间改变位置。例如，刺激后柄2移动更接近PSD。虽然我们以前已经描述了在激活过程中CaMKII向PSD的运动，但我们现在已经表明另一种主要的调节蛋白SynGAP在刺激时远离PSD。刺激诱导的变化与AMPA受体标记在PSD的显着增加相一致，与CaMKII和SynGAP在PSD调节AMPA受体运输中的酶促作用以及结构作用相容。活性诱导的SynGAP易位的机制正在研究中。已知SynGAP与PSD-95结合，这种结合可能是将其定位于PSD核心的锚。在促进CaMKII活化的条件下孵育分离的PSD引起SynGAP的磷酸化并破坏其与PSD-95的免疫共沉淀。 CYLD是另一种在激活后向PSD移动的蛋白质。 CYLD是一种对Lys-63连接的多聚泛素具有特异性的去泛素化酶。如通过蛋白质印迹法和质谱法所见，CYLD高度富集PSD级分。免疫金电镜显示CYLD定位于胞体和树突，但不在轴突，在基础条件下培养的海马神经元中很少PSD标记CYLD。去极化促进CYLD在PSD的进一步积累，使CYLD标记的突触数量加倍，并增加CYLD标记的强度。突触处CYLD活性的增加可以防止突触蛋白的溶酶体降解并调节其细胞内运输。现在已经用针对CYLD的非重叠序列产生的两种抗体证实了PSD处CYLD的富集。 CYLD活性后免疫金标记的增加已被证实并进一步记录。与Ashish Jain，NIAID合作，使用CYLD KO小鼠的新研究比较了蛋白质泛素化的水平和模式以及聚集/洗涤剂不溶性蛋白质的量与野生型小鼠，

 分离的PSD的EM断层扫描被用于更精确地描述蛋白质在PSD处的活动后的重新定位。两项技术突破，从海马切片培养中分离PSD组分，以及与EM断层扫描兼容的负染色方法，是这项新工作的基础。从去极化海马切片分离的PSD显示出升高的CaMK II水平，证实在PSD处的活性诱导的变化在分离期间被保留。CaMKII关联结构域在负染色断层图像中容易识别，从而允许PSD内的个体CaMKII分子的准确定位。我们计划在刺激方案（例如chemLTP和chemLTD）后确切地确定PSD中CaMKII结合的位置，因为CaMKII在PSD内的多个位置处具有多个磷酸化靶标。定量质谱合作到位，以跟踪蛋白质组成和磷酸化的平行变化。

Movements of PSD components during activity were studied by pre-embedding Nanogold labeling of dissociated hippocampal neurons in culture. Depolarization with high K+ is typically used to depolarize and activate neuronal cultures, but other protocols closer to physiological stimulation, such as application of NMDA and synaptic activation with glycine in the absence of Mg2+, are also used. Now that antibodies for nearly all major components of the PSD have been obtained and tested for compatibility for immunoelectron microscopy, a general picture of activity-induced reorganization of the PSD at the molecular level begins to emerge. Scaffold proteins such as PSD-95 and GKAP appear to be relatively stable during short-term activity, and constitute a layer 30 nm thick lying next to the post synaptic membrane where receptors are stabilized by associations with the scaffold. 

In contrast, other proteins change location during short-term activity. For instance, Shank2 moves closer to the PSD after stimulation. While we have previously described movement of CaMKII towards the PSD during activation, we have now showed that another major regulatory protein, SynGAP, moves away from the PSD upon stimulation. The changes induced by stimulation coincide with a marked increase of AMPA receptor labeling at the PSD, compatible with enzymatic as well as structural roles of CaMKII and SynGAP in regulating AMPA receptor trafficking at the PSD. The mechanism of activity-induced SynGAP translocation is being investigated. SynGAP is known to bind to PSD-95 and this association may be the anchor that localizes it to the PSD core. Incubation of isolated PSDs under conditions that promote CaMKII activation causes phosphorylation of SynGAP and disrupts its co-immunopreciptation with PSD-95. CYLD is another protein that moves toward the PSD upon activation. CYLD is a deubiquitinase specific for Lys-63 linked polyubiquitins. CYLD is highly enriched in PSD fractions as seen by western blotting and mass spectrometry. Immunogold EM shows localizes CYLD in cell bodies and dendrites, but not in axons, and few PSDs in cultured hippocampal neurons label for CYLD under basal conditions. Depolarization promotes further accumulation of CYLD at PSDs, doubling the number of synapses labeled for CYLD as well as increasing the intensity of CYLD labeling. Increased CYLD activity at the synapse could prevent lysosomal degradation of synaptic proteins and regulate their intracellular trafficking. Enrichment of CYLD at the PSD has now been confirmed with two antibodies raised to non- overlapping sequences of CYLD. The increase in immunogold labeling for CYLD upon activity has been confirmed and further documented. New studies using CYLD KO mice, a collaboration with Ashish Jain, NIAID, compare levels and patterns of protein ubiquitination and amounts of aggregated/detergent insoluble proteins with wild type mice,

 EM tomography of isolated PSDs is being used to describe with more precision the re-localization of proteins at the PSD following activity. Two technical breakthroughs, the isolation of PSD fractions from hippocampal slice cultures, and a negative staining method compatible with EM tomography, lie behind this new work. PSDs isolated from depolarized hippocampal slices show elevated levels of CaMKII confirming that activity-induced changes at the PSD are preserved during isolation. CaMKII association domains are readily recognizable in negative stain tomograms, allowing accurate localization of individual CaMKII molecules within the PSD. We plan to determine exactly where in the PSD CaMKII is bound after stimulation protocols, such as such as chemLTP and chemLTD, because CaMKII has multiple phosphorylation targets at multiple locations within the PSD. Quantitative mass spectrometry collaborations are in place to track parallel changes in protein composition and phosphorylation.