Abstract:

The advancement of synthetic biology depends on the breakthroughs in the "design", "build", "test", and "learn" (DBTL) stages. With the rapid methodological innovations in genome sequencing, editing, synthesis, and artificial intelligence, the industry has made remarkable progress in designing and building mutants and even artificial cell factories. However, the "unmanageable complexity of large systems" remains one of the ongoing challenges for synthetic biology. As the cell mutant libraries get larger, the process of testing becomes more tedious and even impossible. Hence, there is an urgent need to develop high throughput sorting platforms. SCRS (single-cell Raman spectrum)technology can identify panoramic information at the single-cell level in a non-labeled state, distinguishing complex functional phenotypes. It has advantages such as being fast, low-cost, and capable of being coupled with downstream omics research, thus making it a novel technology for single-cell phenotype identification. At present, based on the powerful ability of SCRS in phenotyping, a series of synthetic phenotypic testing and cell sorting equipment have been developed and a wide range of application demonstrations have been carried out, which demonstrates its enormous potential in accelerating the phenotypic testing and cell sorting in synthetic biology. In this review, we selected the self-developed Raman-activated cell sorting coupled sequencing system (RACS-Seq), single-cell microfluidic droplet sorting system (EasySort), and high-throughput flow cytometry Raman-activated cell sorting (FlowRACS) as the typical equipment, by introducing their technical principles, technical iterations and characteristic application cases. Despite the advances in SCRS-based synthetic phenotypic testing and cell sorting equipment, there are still challenges to overcome. For example, there is a need for improving automation and standardizing protocols to ensure reproducibility and scalability. The development of more powerful artificial intelligence algorithms for dissecting SCRS is also required to exploit more complicated phenometypes. Finally, the throughput and sensitivity still need to be improved significantly. In conclusion, in spite of some disadvantages, the SCRS-based equipment has shown great promise in accelerating the phenotypic testing and cell sorting in synthetic biology.

Key words: synthetic biology, Raman spectroscopy, cell factory, single-cell phenotyping, high-throughput sorting

摘要：

合成生物学的跨越式发展，取决于“设计-构建-测试-学习”（design-build-test-learn）这四大环节的突破。随着基因组测序、编辑、合成以及人工智能技术的日新月异，业界设计和构建突变体甚至人工细胞工厂的能力已经突飞猛进。然而，合成生物学至今仍面临的困境之一便是“大体系的复杂性难以处理”，一旦体系变大，细胞表型测试与分选的工作量就非常艰巨，甚至不可完成。单细胞拉曼光谱（SCRS）技术能够在活体单细胞水平、非标记状态下识别全景信息从而分辨复杂功能表型，且具有快速、低成本、能够与下游细胞组学研究耦联等优势，被视为全新的单细胞表型识别技术。目前，基于SCRS技术强大的表型识别能力已发展了系列合成表型的测试与分选装备，并进行了广泛的应用示范，展示了其助力合成生物学表型测试与分选的巨大潜力。本文选取自主研制的单细胞拉曼光镊分选仪（RACS-Seq）、单细胞微液滴分选系统（EasySort）和高通量流式拉曼分选仪（FlowRACS）为典型仪器装备，分别概述其技术原理和技术迭代以及特色应用案例等。本文最后对当前基于SCRS技术的合成表型测试分选装备所存在的问题及潜在解决策略进行了探讨和展望。

关键词: 合成生物学, 单细胞拉曼光谱技术, 细胞工厂, 单细胞表型识别, 高通量分选