Sainsbury Wellcome Centre for Neural Circuits and Behaviour, University College London, London, UK.
Neuron. 2022 Aug 3;110(15):2470-2483.e7. doi: 10.1016/j.neuron.2022.05.011. Epub 2022 Jun 10.
Processing of sensory information depends on the interactions between hierarchically connected neocortical regions, but it remains unclear how the activity in one area causally influences the activity dynamics in another and how rapidly such interactions change with time. Here, we show that the communication between the primary visual cortex (V1) and high-order visual area LM is context-dependent and surprisingly dynamic over time. By momentarily silencing one area while recording activity in the other, we find that both areas reliably affected changing subpopulations of target neurons within one hundred milliseconds while mice observed a visual stimulus. The influence of LM feedback on V1 responses became even more dynamic when the visual stimuli predicted a reward, causing fast changes in the geometry of V1 population activity and affecting stimulus coding in a context-dependent manner. Therefore, the functional interactions between cortical areas are not static but unfold through rapidly shifting communication subspaces whose dynamics depend on context when processing sensory information.
感觉信息的处理依赖于层次化连接的新皮层区域之间的相互作用，但目前尚不清楚一个区域的活动如何因果地影响另一个区域的活动动态，以及这种相互作用随时间的变化有多快。在这里，我们表明，初级视觉皮层（V1）和高级视觉区域 LM 之间的通信是上下文相关的，并且随着时间的推移令人惊讶地具有动态性。通过在记录另一个区域的活动的同时暂时抑制一个区域，我们发现当老鼠观察视觉刺激时，两个区域都能在一百毫秒内可靠地影响目标神经元的不断变化的亚群。当视觉刺激预测奖励时，LM 反馈对 V1 反应的影响变得更加动态，导致 V1 群体活动的几何形状快速变化，并以依赖上下文的方式影响刺激编码。因此，皮质区域之间的功能相互作用不是静态的，而是通过快速变化的通信子空间展开的，当处理感觉信息时，其动力学取决于上下文。
