Department of Neurobiology and Anatomy, University of Rochester School of Medicine, Rochester, New York 14642, USA.
J Neurophysiol. 2012 Nov;108(9):2452-72. doi: 10.1152/jn.00914.2011. Epub 2012 Aug 15.
Rapidly changing environments require rapid processing from sensory inputs. Varying deflection velocities of a rodent's primary facial vibrissa cause varying temporal neuronal activity profiles within the ventral posteromedial thalamic nucleus. Local neuron populations in a single somatosensory layer 4 barrel transform sparsely coded input into a spike count based on the input's temporal profile. We investigate this transformation by creating a barrel-like hybrid network with whole cell recordings of in vitro neurons from a cortical slice preparation, embedding the biological neuron in the simulated network by presenting virtual synaptic conductances via a conductance clamp. Utilizing the hybrid network, we examine the reciprocal network properties (local excitatory and inhibitory synaptic convergence) and neuronal membrane properties (input resistance) by altering the barrel population response to diverse thalamic input. In the presence of local network input, neurons are more selective to thalamic input timing; this arises from strong feedforward inhibition. Strongly inhibitory (damping) network regimes are more selective to timing and less selective to the magnitude of input but require stronger initial input. Input selectivity relies heavily on the different membrane properties of excitatory and inhibitory neurons. When inhibitory and excitatory neurons had identical membrane properties, the sensitivity of in vitro neurons to temporal vs. magnitude features of input was substantially reduced. Increasing the mean leak conductance of the inhibitory cells decreased the network's temporal sensitivity, whereas increasing excitatory leak conductance enhanced magnitude sensitivity. Local network synapses are essential in shaping thalamic input, and differing membrane properties of functional classes reciprocally modulate this effect.
快速变化的环境需要快速处理感官输入。啮齿动物初级面部触须的偏转速度变化会导致腹后内侧丘脑核内的时间神经元活动模式发生变化。单个体感皮层 4 桶中局部神经元群体根据输入的时间分布将稀疏编码的输入转换为基于输入的尖峰计数。我们通过创建一个类似桶的混合网络来研究这种转换，该网络使用来自皮质切片制备的体外神经元的全细胞记录，通过电导钳通过呈现虚拟突触电导将生物神经元嵌入模拟网络中。利用混合网络，我们通过改变对不同丘脑输入的桶状群体反应来检查互惠网络特性（局部兴奋性和抑制性突触收敛）和神经元膜特性（输入电阻）。在局部网络输入的情况下，神经元对丘脑输入的时间更具选择性；这是由于强前馈抑制。强抑制（阻尼）网络状态对时间更具选择性，对输入的幅度选择性较小，但需要更强的初始输入。输入选择性在很大程度上取决于兴奋性和抑制性神经元的不同膜特性。当抑制性和兴奋性神经元具有相同的膜特性时，体外神经元对输入的时间与幅度特征的敏感性大大降低。增加抑制性细胞的平均泄漏电导会降低网络的时间敏感性，而增加兴奋性泄漏电导会增强幅度敏感性。局部网络突触对于塑造丘脑输入至关重要，而功能类别的不同膜特性会相互调节这种效应。
