Yao Yifan, Pauls Scott, Foley Duncan, Yoshikawa Tomoko, Honma Sato, Honma Ken-Ichi, McVeigh Ellie, Foley Nicolas C, Silver Rae
Department of Psychology, Columbia University, New York, New York, United States of America.
Department of Mathematics, Dartmouth College, Hanover, New Hampshire, United States of America.
PLoS Comput Biol. 2025 Mar 6;21(3):e1012855. doi: 10.1371/journal.pcbi.1012855. eCollection 2025 Mar.
The suprachiasmatic nucleus (SCN), locus of a circadian clock, is a small nucleus of approximately 20,000 neurons that oscillate with a period of about 24 hours. While individual neurons produce circadian oscillations even when dispersed in culture, the coherence and robustness of oscillation of the SCN as a whole is dependent on its circuitry. Surprisingly, the individual neurons of the intact SCN do not all oscillate in phase with each other. To understand the oscillatory dynamics across the intact nucleus, we develop a model of the relation of the phase of neurons to their PER2 expression at a particular subjective time (CT1900) using time series data from SCN slice preparations. Next, we use the model, which produces a surprisingly good fit in the SCN slice data, to estimate oscillator phase at a single time point (CT1900) in snapshot data from PER2 expression measurements in intact, unsliced SCN-wide tissue. To monitor temporal changes in phase in time series data, we use PER2::LUC imaging in an ex vivo SCN slice preparation. To study phase in the intact SCN at a fixed time point we use data generated by PER2 staining and a tissue clearing protocol. Because PER2 expression, as measured in the time series slices and the snapshot intact SCN are not directly comparable, the model estimated from time series slices to the snapshot intact SCN data requires a calibrating constant. The results indicate that our model provides a surprisingly good fit to the SCN slice data and is therefore a meaningful method for estimating phase in the intact SCN snapshot data, permitting the study of virtual interventions such as virtual tissue slicing. We next compare oscillation in circuits in the SCN-wide tissue to those that have been disrupted by virtual slicing using a Kuramoto model to simulate the dynamics. The results support prior evidence that the damage done by coronal slicing has the most disruptive impact on SCN oscillation, while horizontal slicing has the least damage. The results point to the importance of connectivity along the caudal-to-rostral axis and indicate that SCN circuit organization depends on the caudal-to-rostral flow of information. In summary, the construction of this model is a major finding of the paper. Our modeling allows us to perform the previously impossible analysis of oscillatory dynamics in static data in an intact SCN captured at a single time point.
视交叉上核（SCN）是昼夜节律钟的所在位置，它是一个由大约20,000个神经元组成的小核，这些神经元以约24小时的周期振荡。虽然单个神经元即使分散在培养物中也会产生昼夜节律振荡，但整个SCN振荡的连贯性和稳健性取决于其神经回路。令人惊讶的是，完整SCN中的单个神经元并非都彼此同相振荡。为了了解整个完整核中的振荡动力学，我们使用来自SCN切片制备的时间序列数据，建立了一个关于神经元相位与其在特定主观时间（CT1900）的PER2表达之间关系的模型。接下来，我们使用该模型（它在SCN切片数据中给出了惊人的良好拟合）来估计完整的、未切片的全SCN组织中PER2表达测量的快照数据在单个时间点（CT1900）的振荡器相位。为了监测时间序列数据中相位的时间变化，我们在离体SCN切片制备中使用PER2::LUC成像。为了在固定时间点研究完整SCN中的相位，我们使用由PER2染色和组织透明化方案生成的数据。由于在时间序列切片和完整SCN快照中测量的PER2表达不可直接比较，从时间序列切片到完整SCN快照数据估计的模型需要一个校准常数。结果表明，我们的模型与SCN切片数据拟合得非常好，因此是一种用于估计完整SCN快照数据中相位的有意义的方法，允许研究虚拟干预，如虚拟组织切片。接下来，我们使用Kuramoto模型模拟动力学，比较全SCN组织中的回路振荡与那些因虚拟切片而被破坏的回路振荡。结果支持了先前的证据，即冠状切片造成的损伤对SCN振荡具有最大的破坏影响，而水平切片造成的损伤最小。结果表明了沿尾端到头端轴的连接性的重要性，并表明SCN回路组织依赖于从尾端到头端的信息流。总之，该模型的构建是本文的一个主要发现。我们的建模使我们能够对在单个时间点捕获的完整SCN中的静态数据进行以前无法进行的振荡动力学分析。
