Abstract:

As an important technique in biotechnology and bioengineering, animal cell culture often requires frequent analysis and monitoring of some key biomass and biochemical properties. Current practice relying on sampling and off-line analysis inevitably has time delays and potential risk of bacterial contamination. Monitoring the culture process of animal cells on-line in real-time using process analytical technology (PAT), and subsequently achieving process optimization and automatic control, is of vital academic and industrial significance. This review focuses on two most promising PAT techniques for animal cell culture monitoring, i.e., in situ microscopic imaging and Raman spectroscopy. The former characterizes such important properties of biomass as cell number, dimensions, and morphologies, while the latter measures the biochemical indices of concentrations of nutrients, metabolites, and protein products. Emphasis of the review is on methods for image analysis and model building using Raman spectral data, as they hold the key to the successful deployment of the PAT tools. Brief introduction of some other emerging PAT techniques is also given to near infrared spectroscopy, fluorescence spectroscopy, and so forth. Furthermore, built on the advancement of PAT, future perspectives for automatic control and intelligent manufacturing for animal cell culture are explored.

Key words: process analytical technology, cell culture, Raman spectroscopy, image analysis, bioreactors, algorithm, optimization

摘要：

动物细胞培养是一项重要的生物工程技术，在该过程中需要对一些重要的生物和生化指标进行监测。目前广泛采用的是取样离线分析手段，存在时间滞后性和潜在的染菌风险。利用过程分析技术（process analytical technology，PAT）在线实时检测动物细胞培养状况，进而实现培养过程的优化与控制，具有重要的学术意义和工业应用价值。本文介绍了两项应用前景最为广阔的细胞培养PAT技术，即原位显微镜成像与拉曼光谱技术，分别用于在线测量细胞培养过程的重要生物指标（细胞数目、尺寸、形状）与生化指标（营养物质、代谢物质、蛋白质产物的浓度）。作为这两项技术成功应用的关键，文中重点介绍了图像分析算法和拉曼光谱建模方法。此外，对近红外光谱、荧光光谱等技术的一些研究探索也做了简要概括性介绍。最后，结合PAT技术的发展，探讨了实现动物细胞培养过程自动化与智能化的未来发展方向。

关键词: 过程分析技术, 细胞培养, 拉曼光谱, 图像分析, 生物反应器, 算法, 优化