視丘下核單一神經元中多樣的行為表徵與皮質神經輸入的匯聚

Abstract

視丘下核（Subthalamic nucleus, STN）在基底核網絡中控制運動和認知功能扮演至關重要的角色，其神經活動與帕金森氏症等疾病關聯密切。STN的神經活動在臨床的重要性更是體現在對帕金森氏症病人的深腦刺激（Deep-brain stimulation, DBS）治療中。然而，對於單一STN神經元如何編碼不同行為狀態，目前的理解仍然有限。 本篇論文的目的在揭示STN的個別神經元活動以及其神經群體如何在不同情境下編碼行為。透過對小鼠進行活體的鈣離子影像的記錄，我們追蹤了個別STN神經元在多種行為狀況下的活動模式，其中包含自發性跑動和舔舐行為，以及獎勵驅動的反應。研究結果顯示，STN神經元對運動和獎勵驅動的行為反應各異，呈現出豐富的功能多樣性。絕大多數神經元在運動開始時活化，但細胞間的反應在較細微的時間動態上具有差異。而對於獎勵驅動的舔舐行為，它們的神經反應更加地截然不同，其中包含了抑制性及活化性的鈣離子活動。此外，我們發現單個STN神經元能夠表徵多種行為條件，且不同神經元多重表徵的樣式也不同。從神經群體動態學(Neural population dynamics)的角度看，我們發現STN反應中的兩個內在成分能夠分別和跑動速度與舔舐強度高度相關，暗示了這個腦區可能的基本計算原則。另外，透過跨突觸神經路徑的追蹤，我們發現了來自不同皮質區域的輸入在STN具有一定程度的重疊及匯聚，這呼應了單一神經元多樣的行為表徵。 這項研究挑戰了STN主要參與基本運動控制的傳統觀點，並強調它能整合和編碼複雜的行為信息。實驗中所觀察到STN在功能上的異質性顯示了這個核區的功能區域不應僅局限於運動和非運動訊息的空間劃分，實際上可能存在不同功能路徑的重疊與匯聚。這些新見解有望提高STN-DBS治療方法的精準度。

The subthalamic nucleus (STN) plays a critical role in the modulation of motor and cognitive functions within the basal ganglia network, with its activity intricately linked to the pathophysiology of disorders such as Parkinson's disease (PD). This is especially relevant in the context of the widely applied STN deep brain stimulation (DBS) for treating PD patients. However, despite its clinical significance, our understanding of the functions of single STN neurons remains incomplete, particularly regarding how these neurons encode various behavioral states. This thesis aims to elucidate the neural dynamics of the STN at the single-cell level and explore how its neuronal population dynamics encode different behaviors under varied contextual influences. Employing in vivo two-photon calcium imaging with the endoscopic lens in mice, we mapped the activity patterns of individual STN neurons to multiple behavioral conditions, including self-initiated locomotion, licking, and reward-driven responses. Our findings reveal that STN neurons display a rich diversity in response to both motor and reward-driven behaviors, with different subgroups of neurons showing distinct temporal dynamics. Most neurons activate at the onset of movement, each exhibiting nuanced temporal patterns. In contrast, responses to reward-driven licking are uniquely distinct, displaying both inhibition and activation patterns. Furthermore, we demonstrate that individual STN neurons can represent multiple behavioral conditions, with mixed patterns of representation observed across different neurons. From a population dynamics perspective, low-dimensional components extracted from the STN responses were identified to represent locomotion speed and licking intensity, implying potential computational basic principles in this structure. Finally, anterograde transsynaptic labeling targeting the hyper-direct pathway of the basal ganglia illustrates the convergence and overlap of cortical inputs in the STN, which corresponds to the diverse behavioral representations of individual neurons. The study challenges the traditional view that the STN is primarily involved in basic motor control, instead highlighting its capacity to integrate and encode complex sets of behavioral information. Moreover, the observed heterogeneity suggests that the STN's functional territories are not strictly segregated into motor and non-motor spatial domains as previously thought. Instead, there is likely overlap and convergence of distinct functional pathways. These insights have the potential to enhance the precision of STN-DBS targeting for treating Parkinson's disease.