Synaptic Inputs to Retinal Ganglion Cells

描述（申请人提供）：哺乳动物视网膜的神经节细胞大约有十几种类型，每一种都向大脑传递不同的视觉场景编码。这项研究的长期目标是了解特定类型的神经节细胞实现这些融合的机制。我们开发了一种界面培养系统，可以在无监督的培养系统中维持多个成年兔视网膜样本数天。编码RNAi或标记突触蛋白的基因被生物转染到单个视网膜神经节细胞中。

DESCRIPTION (provided by applicant): The ganglion cells of mammalian retinas exist in roughly a dozen types, each transmitting a different encoding of the visual scene to the brain. The long-term goal of this research is to learn the mechanism by which specific types of ganglion cells achieve these codlings. We have developed an interface incubation system that allows maintenance of multiple samples of adult rabbit retinas for several days in an unsupervised, culture-like system. Genes coding for RNAi or tagged synapse proteins are biolistically transfected into individual retinal ganglion cells. Two questions involve the synaptic events underlying directional selectivity in certain retinal ganglion cells. First, we propose to use RNAi to knock down GABAergic or cholinergic responsiveness in individual ganglion cells. Recording will reveal the contribution of these direct (postsynaptic) inputs to direction selectivity. Second, we will investigate the co-release of GABA and Ach by the starburst amacrine cells. Are the two neurotransmitters released from the same cellular sites or different ones? This will be studied by localizing their vesicular transport proteins. The third question is a more general one. Our recent experiments suggest that the excitatory inputs to ganglion cell dendrites are spatially distributed according to an even-spacing law: the synapses seem to repel each other. In addition, they systematically avoid branch points in the dendritic arbor. We now propose to see if these rules apply to all types of ganglion cells; to model their physiological consequences; and to see if the same rules apply to inhibitory synapses. The methods introduced here can be used in adult animals of any species. They obviate the need for transgenic animals, increase the number of samples that can be simultaneously studied, and allow labeling of proteins that have long half-lives. They can be used in normal tissue or in disease models. We hope that they will be useful to laboratories studying both basic and clinical problems.