Model Neuron Encoding the Realistic Learning Rule

描述(由申请人提供)：活动依赖的突触修饰、长时程增强(LTP)和抑制(LTD)在神经网络的信息处理和存储中是必不可少的。因此，了解突触改变的机制对于理解学习和记忆功能至关重要。我们研究的长期目标是了解突触修饰如何参与海马神经网络的信息处理。然而，中枢神经网络的复杂性阻碍了仅通过生物学方法解决这些问题。因此，我们建议结合电生理学和计算建模方法来开发一个模型神经元系统，该系统编码由大鼠海马片突触前和突触后相关活动引起的突触修改。这些研究将使我们能够预测突触修饰如何调制神经网络中超出当前电生理技术的多个时空不同的输入。 我们先前已经通过对海马CA1神经元的电生理研究证明：1)5 Hz的相关活动可以诱导LTP或LTD，这取决于突触前和突触后激活的精确时间；2)激活部位LTP的窄时间窗(15ms)两侧是LTD的两个时间窗，可以从刺激(同突)部位扩展到非刺激(异合)部位；3)LTP和LTD之间的转换发生在25ms内，这是40 Hz振荡的特征时间。此外，突触后钙离子通过N-甲基-D-天冬氨酸受体的钙离子内流，以及通过兰尼定和IP3受体从内库差异释放的钙离子，调节活性诱导的突触修饰的极性和输入特异性。我们的发现表明，依赖活动的突触修饰和海马区定位细胞的振荡模式之间存在联系，这被认为是负责空间记忆的。 我们采用电生理学、钙离子和光学成像相结合的综合方法，并在海马片上进行计算建模，旨在解决以下问题：1)突触可塑性从激活的突触传播到非激活的突触的时空模式是什么？我们将通过逆行信号描述LTP和LTD在突触后沿树突或突触前的传播。2)抑制性神经元如何参与神经网络中突触修改的极性和程度的确定？3)突触后神经元中如何整合不同路径上基于峰时的突触修改？利用基于棘波时序的诱导协议，我们将研究来自具有不同时空差异的输入的信号如何整合到单个突触后神经元中。这些研究将有助于我们理解学习和记忆的细胞和分子基础。

DESCRIPTION (provided by applicant): Activity-dependent synaptic modifications, long-term potentiation (LTP) and depression (LTD) are essential in information processing and storage in neural networks. Thus, understanding mechanisms of synaptic modifications are crucial in understanding learning and memory functions. The long-term goal of our research is to understand how synaptic modifications are involved in information processing in hippocampal neural network. However, the complexity of central neural networks hinders addressing these issues through biological approaches alone. Therefore, we propose to combine electrophysiologic and computational modeling approaches to develop a model neuron system encoding synaptic modifications induced by correlated pre- and postsynaptic activity in rat hippocampal slices. These studies will enable us to predict how synaptic modifications modulate the multiple spatiotemporal-distinct inputs in neuronal networks that is beyond current electrophysiological techniques. Previously we have demonstrated by electrophysiological studies in hippocampal "CAI neurons that 1) correlated activity at 5 Hz can induce either LTP or LTD, depending on the precise timing of pre-and postsynaptic activation, 2) a narrow time window (15 ms) that exists for LTP of the activated site is flanked by two time windows for LTD, which can spread from stimulated (homosynaptic) to non-stimulated (heterosynaptic) sites, 3) the transition between LTP and LTD occurs within 25 ms, a characteristic time for 40 Hz oscillations. Furthermore, the postsynaptic Ca2+, derived from Ca2+ influx via N-methyl-D-aspartate receptors and a differential release of Ca2+ from internal stores via ryanodine and IP3 receptors, regulates both polarity and input specificity of activity-induced synaptic modificalion. Our findings suggest a link between activity-dependent synaptic modifications and oscillation patters of place cells in the hippocampus, which are believed to be responsible for spatial memory. Using an integrative approach that combines electrophysiology, Ca2+ and optical imaging, and computational modeling in hippocampal slices, we specifically aims to address following questions: 1) What is the spatiotemporal pattern of spread of synaptic plasticity from activated synapses to non-activated synapses? We will characterize the spread of LTP and LTD postsynaptically along the dendritic arbor or presynaptically via retrograde signaling. 2) How are inhibitory neurons involved in determination of polarity and extent of synaptic modifications in neural networks? 3) How are spike-timing based synaptic modifications at different pathways integrated in a postsynaptic neuron? Taking advantage of the spike-timing based induction protocol, we will investigate how signals originated from inputs with various spatiotemporal differences can be integrated in a single postsynaptic neuron. These studies will contribute significantly to our understanding the cellular and molecular basis of learning and memory.