小清蛋白中间神经元的成熟在发育中重塑皮层群体活动结构
Parvalbumin interneuron maturation reorganizes cortical population activity structure during development
在大脑发育过程中，自发皮层活动经历了一场剧烈的转变：从高度同步的网络事件转变为相关性更低、更稀疏的神经元放电，这一转变被认为构成了更灵活的神经编码的基础。然而，这一转变过程中群体活动的结构如何变化、哪些环路机制驱动了这一重组，以及其对信息编码的意义，尚未被完全确立。利用对出生后早期小鼠的纵向在体钙成像，我们证明，在出生后第二周，皮层网络去同步化伴随着分数方差从占主导的低维模式重新分配到许多较弱的维度，以及GABA能中间神经元（INs）对网络的影响不断增强。全光学环路探查揭示，小清蛋白（PV）INs——而非生长抑素（SST）INs——足以使网络去同步化并改变群体方差的分布。通过对一部分PV INs进行靶向消融，我们表明维度与刺激解码准确度高度相关。此外，PV INs在出生后第8天（P8）至第11天之间经历转录成熟的急剧增加，在此期间它们优先富集突触相关基因。最后，一个新颖的、生物学上合理的多细胞类型计算模型再现了发育性去同步化和维度的上升。引人注目的是，在不受扰动数据约束的情况下，该模型准确预测了我们所观察到的每一类IN对网络的影响。
用纵向钙成像和全光学操纵把 PV 中间神经元成熟与群体活动维度升高联系起来，并给出可预测扰动结果的计算模型。
During brain development, spontaneous cortical activity undergoes a dramatic transformation from highly synchronous network events to less correlated and sparse neuronal firing that is thought to underlie more flexible neural coding. However, how the structure of population activity changes across this transition, what circuit mechanisms drive this reorganization, and the implications for information encoding have not been fully established. Using longitudinal in vivo calcium imaging in early postnatal mice, we demonstrate that cortical network desynchronization during the second postnatal week is accompanied by a redistribution of fractional variance from dominant, low-dimensional modes into many weaker dimensions, and a growing influence of GABAergic interneurons (INs) on the network. All-optical circuit interrogation revealed that parvalbumin (PV) INs, but not somatostatin (SST) INs, are sufficient to desynchronize the network and shift the distribution of population variance. Through targeted ablation of a fraction of PV INs, we show dimensionality is highly correlated with stimulus decoding accuracy. Furthermore, PV INs undergo a sharp increase in transcriptional maturation between postnatal day (P) 8 and 11, during which they preferentially enrich for synaptic genes. Lastly, a novel, biologically plausible, multi-cell-type computational model reproduced the developmental desynchronization and rise in dimensionality. Strikingly, without being constrained by perturbation data, the model accurately predicted the effects of each IN class on the network that we observed empirically.
中文翻译还没有生成，先显示原文。
来源：bioRxiv · Neuroscience · biorxiv.org