Characterizing and modulating motor cortical dynamics underlying rapid sequence learning in primates

项目摘要 一个根本性的重要问题是神经系统如何收敛到最优解， 运动学习的过程。越来越多的人类证据已经显著地证明了 微离线增益（MOG），或短暂休息（约10秒）后的显著“离线”性能增益， 在运动序列学习过程中，随着快速和可靠的性能得到巩固，这种能力会减弱。 脑磁图（MEG）记录将这种快速形式的巩固与13-30 Hz的振荡联系起来 在场电位（β，β）和重放的宽带场电位模式在运动皮层，特别是 初级运动皮层（M1）--运动技能的执行和学习所必需的区域。但 尚不清楚M1中的高分辨率尖峰信号如何反映在空间宽MEG的再激活中 信号以及离线β可能如何支持整合。此外，目前还不清楚这种微离线处理是否 导致快速的行为改变 在这里，我们使用了一种新的非人灵长类动物（NHP）的顺序到达任务，可靠地elaims MOG，结合 与LFP和运动皮层神经元尖峰记录，以探讨灵长类动物运动皮层如何使快速 序列学习我们的初步数据表明，M1中的任务激活神经元系综在 短暂的休息，特别是在MOG最高的早期学习中。相比之下，离线β在后期期间最高 断裂，并与MOG呈负相关。总之，这些结果激发了我们的总体假设， 任务活跃的尖峰模式的微离线重新激活促进了快速学习， 优化的离线β增加，以促进学习的神经活动模式的稳定性。 为了验证这一假设，我们使用了跨学科的方法，高速到达和凝视跟踪，精确 神经记录、计算建模和因果操纵。在目标1中，我们将评估 在短暂休息期间任务活跃的集合的重新激活与快速的行为改变相关。在 目标2，我们将量化离线β-相干尖峰模式和在线尖峰变化之间的关系 动力学最后，在目标3中，我们将使用20 Hz交流电刺激（ACS）来因果地确定 离线β在调节快速巩固和行为稳定性中的作用。这些实验将进一步 我们对灵长类动物皮层如何在短时间内实现适应性行为修改的理解， 为基于刺激的行为和认知病理状况干预奠定坚实的基础 僵硬，如帕金森病，强迫症和抑郁症。

PROJECT SUMMARY A fundamentally important question is how the nervous system converges on optimal solutions during the process of motor learning. A growing body of evidence in humans has remarkably demonstrated the presence of micro-offline gains (MOGs), or significant “offline” performance gains after a brief rest period (~10 second), during motor sequence learning, which diminish as a fast and reliable performance is consolidated. Magnetoencephalography (MEG) recordings have linked this rapid form of consolidation to 13-30 Hz oscillations in field potentials (β, beta) and replay of broad-band field potential patterns across the motor cortex, particularly the primary motor cortex (M1) – an area essential to the execution and learning of motor skills. However, it remains unclear how high-resolution spiking signals in M1 are reflected in reactivations of spatially broad MEG signals and how offline β may support consolidation. Moreover, it is unclear if such micro-offline processing is causal to rapid behavioral modifications. Here, we use a novel sequential reach task for non-human primates (NHPs) that reliably elicits MOGs, combined with LFP and neuronal spiking recordings in motor cortex, to probe how the primate motor cortex enables rapid sequence learning. Our preliminary data shows that task-active neuronal ensembles in M1 are reactivated during short breaks, particularly in early learning when MOGs are highest. In contrast, offline β is highest during later breaks and is inversely correlated with MOGs. Together, these results motivate our overall hypothesis that micro-offline reactivation of task-active spiking patterns promotes rapid learning, and as behavior is optimized, offline β increases to promote stability of learned neural activity patterns. To test this hypothesis, we use an interdisciplinary approach of high-speed reach and gaze tracking, precise neural recordings, computational modeling, and causal manipulations. In Aim 1, we will assess whether reactivations of task-active ensembles during brief rest periods correlate with rapid behavioral modifications. In Aim 2, we will quantify the relationship between offline β-coherent spiking patterns and changes in online spiking dynamics. Finally, in Aim 3, we will use 20 Hz alternating current stimulation (ACS) to causally determine the role of offline β in regulating rapid consolidation and behavioral stability. Together, these experiments will further our understanding of how the primate cortex enables adaptive behavioral modifications on short timescales and lay a strong foundation for stimulation-based interventions for pathological conditions of behavioral and cognitive rigidity, such as Parkinson’s disease, obsessive-compulsive disorder, and depression.