Nanoparticle Induced Cell Magneto-Rotation for the Multiplexed Monitoring of Morphology, Stress and Drug Sensitivity of Suspended Single Cancer Cells.

Cancer diagnostic;Metastasis;Clustering and classification using machine learning algorithms;Magneto-rotation;Magnetic nanoparticles;Single cell trapping;Physics;Science;Applied Physics

Elbez, Remy A.Clarke, Roy ;

University of Michigan

关键词: Cancer diagnostic;    Metastasis;    Clustering and classification using machine learning algorithms;    Magneto-rotation;    Magnetic nanoparticles;    Single cell trapping;    Physics;    Science;    Applied Physics;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/111434/relbez_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

The metastatic process of a cancer relies on the transformation of some of the primary tumor cells into cells capable of migrating through the Extra-Cellular Matrix (ECM), surrounding the tumor, into the bloodstream and the lymph nodes, and then settle in distant tissue, growing new secondary tumors. By identifying, characterizing and quantifying these cells, the progression of cancer in a patient during therapy can be more accurately assessed. Here we describe the development of a new method for quantitative real time monitoring of cell size and morphology, on single live suspended cancer cells, unconfined in three dimensions. The enabling cell magnetorotation (CM) method is made possible by nanoparticle induced cell magnetization. Using a rotating magnetic field, the magnetically labeled cells are actively rotated, then imaged, using a high definition CCD camera. Under proper conditions, the rotation period of a magnetic object is proportional to its shape factor. We demonstrate first that the rotational period, when measured in real-time, can serve to track cellular response to drugs, cytotoxic agents and other chemical stimuli. In addition, while cells are rotated, they exhibit very specific morphological activities, even without a chemical stimulus.Described also is how to multiplex the CM method, to image several dozens to several thousands of cells simultaneously, and using morphology to classify cells into different phenotypic categories, with each phenotype being correlated with malignancy level. The intrinsic tumor heterogeneity, at the cellular level, can be visualized with relationship graphs. Shown is the ability to monitor cell morphological changes over long periods of time, in real time, in order to detect the metastatic potential for heterogeneous populations of cancer cells, using tools from statistical analysis methods. The method relies on unsupervised Machine Learning algorithms which do not require human inputs. Overall it is demonstrated that the CM method can be used as a diagnostic tool to evaluate the phenotypical heterogeneity in a cell population in general, and in a cancer cell population in particular. This fast and high throughput method promises to efficiently assess the efficacy of personalized therapeutic strategies.