Constantinidis Christos, Goldman-Rakic Patricia S
Section of Neurobiology, Yale School of Medicine, New Haven, Connecticut 06510, USA.
J Neurophysiol. 2002 Dec;88(6):3487-97. doi: 10.1152/jn.00188.2002.
Neurophysiological recordings have revealed that the discharges of nearby cortical cells are positively correlated in time scales that range from millisecond synchronization of action potentials to much slower firing rate co-variations, evident in rates averaged over hundreds of milliseconds. The presence of correlated firing can offer insights into the patterns of connectivity between neurons; however, few models of population coding have taken account of the neuronal diversity present in cerebral cortex, notably a distinction between inhibitory and excitatory cells. We addressed this question in the monkey dorsolateral prefrontal cortex by recording neuronal activity from multiple micro-electrodes, typically spaced 0.2-0.3 mm apart. Putative excitatory and inhibitory neurons were distinguished based on their action potential waveform and baseline discharge rate. We tested each pair of simultaneously recorded neurons for presence of significant cross-correlation peaks and measured the correlation of their averaged firing rates in successive trials. When observed, cross-correlation peaks were centered at time 0, indicating synchronous firing consistent with two neurons receiving common input. Discharges in pairs of putative inhibitory interneurons were found to be significantly more strongly correlated than in pairs of putative excitatory cells. The degree of correlated firing was also higher for neurons with similar spatial receptive fields and neurons active in the same epochs of the behavioral task. These factors were important in predicting the strength of both short time scale (<5 ms) correlations and of trial-to-trial discharge rate covariations. Correlated firing was only marginally accounted for by motor and behavioral variations between trials. Our findings suggest that nearby inhibitory neurons are more tightly synchronized than excitatory ones and account for much of the correlated discharges commonly observed in undifferentiated cortical networks. In contrast, the discharge of pyramidal neurons, the sole projection cells of the cerebral cortex, appears largely independent, suggesting that correlated firing may be a property confined within local circuits and only to a lesser degree propagated to distant cortical areas and modules.
神经生理学记录显示,附近皮质细胞的放电在时间尺度上呈正相关,其范围从动作电位的毫秒级同步到慢得多的放电率共同变化,这在数百毫秒的平均放电率中很明显。相关放电的存在有助于深入了解神经元之间的连接模式;然而,很少有群体编码模型考虑到大脑皮层中存在的神经元多样性,特别是抑制性和兴奋性细胞之间的区别。我们通过从多个微电极记录神经元活动来解决猴子背外侧前额叶皮层中的这个问题,这些微电极通常相隔0.2 - 0.3毫米。根据动作电位波形和基线放电率来区分假定的兴奋性和抑制性神经元。我们测试了每对同时记录的神经元是否存在显著的交叉相关峰值,并测量了它们在连续试验中的平均放电率的相关性。当观察到交叉相关峰值时,其中心位于时间0,表明同步放电与两个神经元接收共同输入一致。发现假定的抑制性中间神经元对之间的放电比假定的兴奋性细胞对之间的放电显著更强烈相关。对于具有相似空间感受野的神经元以及在行为任务的同一时期活跃的神经元,相关放电的程度也更高。这些因素对于预测短时间尺度(<5毫秒)相关性以及试验间放电率共同变化的强度很重要。试验间的运动和行为变化仅略微解释了相关放电。我们的研究结果表明,附近的抑制性神经元比兴奋性神经元更紧密地同步,并且是未分化皮质网络中常见的大部分相关放电的原因。相比之下,大脑皮层唯一的投射细胞——锥体神经元的放电在很大程度上似乎是独立的,这表明相关放电可能是一种局限于局部回路的特性,并且仅在较小程度上传播到远处的皮质区域和模块。