Pillar of Engineering Product Development, Singapore University of Technology and Design, 8 Somapah Road, 487372 Singapore.
Anal Chem. 2021 Mar 16;93(10):4567-4575. doi: 10.1021/acs.analchem.0c05009. Epub 2021 Mar 4.
Cellular mechanical phenotypes in connection to physiological and pathological states of cells have become a promising intrinsic biomarker for label-free cell analysis in various biological research and medical diagnostics. In this work, we present a microfluidic system capable of high-throughput cellular mechanical phenotyping based on a rapid single-cell hydrodynamic stretching in a continuous viscoelastic fluid flow. Randomly introduced single cells are first aligned into a single streamline in viscoelastic fluids before being guided to a flow splitting junction for consistent hydrodynamic stretching. The arrival of individual cells prior to the flow splitting junction can be detected by an electrical sensing unit, which produces a triggering signal to activate a high-speed camera for on-demand imaging of the cell motion and deformation through the flow splitting junction. Cellular mechanical phenotypes, including cell size and cell deformability, are extracted from the analysis of these captured single-cell images. We have evaluated the sensitivity of the developed microfluidic mechanical phenotyping system by measuring the synthesized hydrogel microbeads with known Young's modulus. With this microfluidic cellular mechanical phenotyping system, we have revealed the statistical difference in the deformability of microfilament disrupted, normal, and fixed NIH 3T3 fibroblast cells. Furthermore, with the implementation of a machine-learning-based classification of MCF-10A and MDA-MB-231 mixtures, our label-free cellular phenotyping system has achieved a comparable cell analysis accuracy (0.9:1, 5.03:1) with respect to the fluorescence-based flow cytometry results (0.97:1, 5.33:1). The presented microfluidic mechanical phenotyping technique will open new avenues for high-throughput and label-free single-cell analysis in diverse biomedical applications.
细胞力学表型与细胞的生理和病理状态有关，已成为各种生物研究和医学诊断中无标记细胞分析的有前途的内在生物标志物。在这项工作中，我们提出了一种基于在连续粘弹性流体流动中快速单细胞流体动力学拉伸的高通量细胞力学表型分析的微流控系统。在被引导到用于一致的流体动力学拉伸的流分裂连接之前，随机引入的单细胞首先在粘弹性流体中被排列成单个流线。在流分裂连接之前到达的单个细胞可以通过电感测单元检测到，该单元产生触发信号以激活高速相机，以通过流分裂连接按需对细胞运动和变形进行成像。从这些捕获的单细胞图像的分析中提取细胞力学表型，包括细胞大小和细胞可变形性。我们通过测量具有已知杨氏模量的合成水凝胶微球来评估所开发的微流控力学表型系统的灵敏度。使用这种微流控细胞力学表型系统，我们揭示了微丝破坏、正常和固定的 NIH 3T3 成纤维细胞的变形能力的统计学差异。此外，通过实施 MCF-10A 和 MDA-MB-231 混合物的基于机器学习的分类，我们的无标记细胞表型系统在细胞分析准确性方面与基于荧光的流式细胞术结果相当（0.9:1, 5.03:1）。所提出的微流控力学表型技术将为各种生物医学应用中的高通量和无标记单细胞分析开辟新途径。
