美国加州大学Lin Tian和霍华德·休斯医学院Loren L. Looger研究组合作取得最新进展。他们通过机器学习指导选择性敏感的血清素传感器的进化。这一研究成果于2020年12月16日发表在国际顶尖学术期刊《细胞》杂志上。

在机器学习的指导下，他们开发并应用了绑定袋重新设计策略，以创建高性能的可溶性荧光5-羟色胺传感器（iSeroSnFR），从而可以光学检测毫秒级的5-羟色胺瞬变。他们证明，iSeroSnFR可用于检测恐惧条件、社交互动和睡眠/觉醒过渡过程中行为自由的小鼠中血清素释放。他们还开发了一种5-羟色胺转运蛋白功能和药物调节的可靠方法。他们希望机器学习指导的绑定口袋重新设计和iSeroSnFR分别在开发其他传感器，以及体外和体内血清素检测方面具有广泛用途。

据了解，5-羟色胺在认知中起着核心作用，并且是大多数精神疾病药物的靶标。现有药物疗效有限。创建改进版本将需要更好地了解5-羟色胺能回路，这已因人们无法以高时空分辨率监测5-羟色胺的释放和运输而受到阻碍。

附：英文原文

Title: Directed Evolution of a Selective and Sensitive Serotonin Sensor via Machine Learning

Author: Elizabeth K. Unger, Jacob P. Keller, Michael Altermatt, Ruqiang Liang, Aya Matsui, Chunyang Dong, Olivia J. Hon, Zi Yao, Junqing Sun, Samba Banala, Meghan E. Flanigan, David A. Jaffe, Samantha Hartanto, Jane Carlen, Grace O. Mizuno, Phillip M. Borden, Amol V. Shivange, Lindsay P. Cameron, Steffen Sinning, Suzanne M. Underhill, David E. Olson, Susan G. Amara, Duncan Temple Lang, Gary Rudnick, Jonathan S. Marvin, Luke D. Lavis, Henry A. Lester, Veronica A. Alvarez, Andrew J. Fisher, Jennifer A. Prescher, Thomas L. Kash, Vladimir Yarov-Yarovoy, Viviana Gradinaru, Loren L. Looger, Lin Tian

Issue&Volume: 2020-12-16

Abstract: Serotonin plays a central role in cognition and is the target of most pharmaceuticalsfor psychiatric disorders. Existing drugs have limited efficacy; creation of improvedversions will require better understanding of serotonergic circuitry, which has beenhampered by our inability to monitor serotonin release and transport with high spatialand temporal resolution. We developed and applied a binding-pocket redesign strategy,guided by machine learning, to create a high-performance, soluble, fluorescent serotoninsensor (iSeroSnFR), enabling optical detection of millisecond-scale serotonin transients.We demonstrate that iSeroSnFR can be used to detect serotonin release in freely behavingmice during fear conditioning, social interaction, and sleep/wake transitions. Wealso developed a robust assay of serotonin transporter function and modulation bydrugs. We expect that both machine-learning-guided binding-pocket redesign and iSeroSnFRwill have broad utility for the development of other sensors and in vitro and in vivo serotonin detection, respectively.

DOI: 10.1016/j.cell.2020.11.040

Source: https://www.cell.com/cell/fulltext/S0092-8674(20)31612-3