在病理学、临床诊断、克隆和细胞生物学研究中，从大量的细胞群体中分离出小量的特殊种类细胞是一项很常用的技术。现较为成熟的方法主要是荧光激发细胞分离法和激光捕获微分法，但都存在各自的不足及局限。
来自麻省理工大学的Voldman等人利用微流细胞分离仪提出了一个简便快捷、基于图像识别的细胞分离方法。微流细胞分离仪装有微孔列阵，能够通过沉降作用来装载哺乳细胞，这一过程可被特制显微镜所探测。随后，利用聚焦红外激光的散射力使所要的细胞从所在的微孔中悬浮起来，进入到一个流动场中而离开列阵表面，这样就可以达到将其收集的目的。
该方法提供了一个界面友好、操作简单、方便快捷的实验平台，可以分离任何在特制显微镜下有外观特性的细胞，并且能和荧光技术很好的结合使用，应用范围十分广泛。我们相信，该项技术成熟化、商品化之后，将对细胞生物学的发展产生重大的影响。 （科学时报 邵振宇/编译）
原文链接：http://pubs.acs.org/cgi-bin/abstract.cgi/ancham/2007/79/i24/abs/ac071366y.html
Anal. Chem., 79 (24), 9321 -9330, 2007. 10.1021/ac071366y S0003-2700(07)01366-2
Web Release Date: November 16, 2007 Copyright © 2007 American Chemical Society
Intuitive, Image-Based Cell Sorting Using Optofluidic Cell Sorting
J. R. Kovac and J. Voldman*
Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Building 36-824, Cambridge, Massachusetts 02139
Received for review June 27, 2007. Accepted September 27, 2007.
Abstract:
We present a microfluidic cell-sorting device which augments microscopy with the capability to perform facile image-based cell sorting. This combination enables intuitive, complex phenotype sorting based on spatio-temporal fluorescence or cell morphology. The microfluidic device contains a microwell array that can be passively loaded with mammalian cells via sedimentation and can be subsequently inspected with microscopy. After inspection, we use the scattering force from a focused infrared laser to levitate cells of interest from their wells into a flow field for collection. First, we demonstrate image-based sorting predicated on whole-cell fluorescence, which could enable sorting based on temporal whole-cell fluorescence behavior. Second, we demonstrate image-based sorting predicated on fluorescence localization (nuclear vs whole-cell fluorescence), highlighting the capability of our approach to sort based on imaged subcellular events, such as localized protein expression or translocation events. We achieve postsort purities up to 89% and up to 155-fold enrichment of target cells. Optical manipulation literature and a direct cell viability assay suggest that cells remain viable after using our technique. The architecture is highly scalable and supports over 10 000 individually addressable trap sites. Our approach enables sorting of significant populations based on subcellular spatio-temporal information, which is difficult or impossible with existing widespread sorting technologies.
