High-throughput deformability-based microfluidic sorting reveals metastatic cancer subpopulations

作者：Xiaoyan Liu, Yuanchao Liu, Junwei Li, Z. Wang, Mui Hoon Nai, Qian Chen, Haodong Li, Xingyu Jiang, Shaofei Shen, Chwee Teck Lim · 发表于：Proceedings of the National Academy of Sciences · 年份：2026 · DOI：10.1073/pnas.2604094123 · 被引用次数：1 · 研究领域：Cancer Cells and Metastasis、Cellular Mechanics and Interactions、3D Printing in Biomedical Research

The deformability of cancer cells is a critical indicator of their malignancy, as this mechanical property of cancer cells can reflect various biostructural and biochemical changes such as the cytoskeletons and protein expressions. However, selectively isolating and analyzing cell subpopulations with differing deformability and their associated biological properties remains challenging, especially at high throughput. Here, we introduce a microstructure-assisted spiral microfluidic platform that sorts cancer cells based on their deformability at ultrahigh throughput (>2,000,000 cells min −1 ) and couples the workflow to real-time, image-based phenotyping of cell parameters. We profile the sorted fractions and investigate a tripartite correlation between the deformability of cancer cells, the expression of proteins in different subpopulations, and the metastatic ability of cancer cells. This tripartite correlation has been confirmed in human breast cancer cell lines with metastatic potential (including MDA-MB-231 cells and BT-549 cells), where these cells are classified into subpopulations based on their deformability. The more deformable subpopulation exhibits increased invasiveness and distinct cytoskeletal remodeling and epithelial-to-mesenchymal transition (EMT)-associated protein signatures. Leveraging these deformability differences, our label-free platform enables high-throughput enrichment of aggressive cancer subpopulations and provides a scalable front end for liqu...