Savaglio Mario Alexios, Brozi Christina, Psilou Eleftheria, Koumpouzi Chryssalenia, Papadostefanaki Marianna, Vasilakos Christos, Nessis Spyros, Smyrnakis Emmanouel, Akoutas Vasileios, Keliris Georgios A, Palagina Ganna, Smirnakis Stelios M, Papadopouli Maria
Department of Computer Science, University of Crete, Greece.
Institute of Computer Science, Foundation for Research and Technology-Hellas, Greece.
bioRxiv. 2025 Sep 1:2025.08.27.672292. doi: 10.1101/2025.08.27.672292.
Variability in single neuron responses presents a challenge in establishing reliable representations of visual stimuli essential for driving behavior. To enhance accuracy, integration of responses from multiple neurons is imperative. This study leverages simultaneous recordings from a large population (tens of hundreds) of neurons, achieved through in vivo mesoscopic 2-photon calcium imaging of the primary visual cortex (V1) in mice, under visual stimulus conditions as well as in resting state (absence of stimulus). The visual stimulus consisted of 16 distinct randomly shuffled directions of motion presented to the mice. We employed mutual information to identify neurons that contain the most significant information about the stimulus direction. As expected, neurons displaying high predictive power (HPP) in stimulus decoding exhibit elevated firing event rates during stimulus presentation. Furthermore, functional connectivity among HPP neurons during visual stimulation is denser and stronger compared to functional connectivity among other visually responsive neurons. Functional connections among HPP neurons appear to form independently of distance, suggesting a distributed yet highly coordinated network. In contrast, HPP neuronal activity and functional connectivity differed significantly at resting state. Specifically, during the resting state, HPP neurons exhibited lower event rates and functional connectivity structure that was not significantly different from that of other visually responsive neurons. This suggests that HPP neurons are less susceptible to being driven simultaneously by internal brain states in the absence of a stimulus. Finally, the tuning properties of HPP neurons were unexpectedly diverse: while some were sharply tuned, others conveyed a similar amount of mutual information, despite exhibiting much weaker tuning. This study sheds light on the organization of neuronal ensembles important for decoding visual motion direction in mouse area V1, contributing to the understanding of information processing in mouse visual cortex.
单个神经元反应的变异性给建立驱动行为所必需的视觉刺激的可靠表征带来了挑战。为了提高准确性，整合多个神经元的反应势在必行。本研究利用通过对小鼠初级视觉皮层（V1）进行体内介观双光子钙成像，在视觉刺激条件以及静息状态（无刺激）下，对大量（数百个）神经元进行同步记录。视觉刺激由呈现给小鼠的16个不同的随机打乱的运动方向组成。我们采用互信息来识别包含有关刺激方向最重要信息的神经元。正如预期的那样，在刺激解码中显示出高预测能力（HPP）的神经元在刺激呈现期间表现出较高的放电事件率。此外，与其他视觉反应神经元之间的功能连接相比，视觉刺激期间HPP神经元之间的功能连接更密集、更强。HPP神经元之间的功能连接似乎独立于距离形成，表明存在一个分布式但高度协调的网络。相比之下，HPP神经元活动和功能连接在静息状态下有显著差异。具体而言，在静息状态下，HPP神经元表现出较低的事件率和功能连接结构，与其他视觉反应神经元的结构没有显著差异。这表明HPP神经元在没有刺激的情况下不太容易受到内部脑状态的同时驱动。最后，HPP神经元的调谐特性出人意料地多样：虽然有些神经元调谐尖锐，但其他神经元尽管调谐弱得多，却传达了相似数量的互信息。这项研究揭示了对解码小鼠V1区视觉运动方向很重要的神经元集合的组织，有助于理解小鼠视觉皮层中的信息处理。
