Distinguishing cancerous from non-malignant breast cells using viscoelastic creep and functional principal component...

The mechanical responses and properties of breast epithelial cells are known to change during malignant transformation and progression due to the dynamics of their actin cytoskeleton network organization and the resulting viscoelastic deformability. Studying the viscoelastic creep behavior of breast cells may reveal new avenues for developing novel cancer diagnostic and therapeutic biomarkers and improving fundamental biophysical understanding of the disease. Here we present an approach that uses functional principal component analysis (fPCA) to distinguish between the viscoelastic responses of malignant and non-malignant live breast cells that are subjected to shear flow in microfluidic channels under in-situ observation with optical, fluorescence, and confocal microscopy. The fPCA method extracts critical features of cell viscoelasticity from the in-situ measured creep responses of non-tumorigenic breast cells (MCF-10A), less metastatic triple-negative breast cancer (TNBC) cells (MDA-MB-468), and highly metastatic breast cancer cells (MDA-MB-231). The results demonstrate distinguishable clustering patterns for the three types of cells in the first principal component (PC) and the second PC space. The first PC, indicative of the overall level of creep compliance, accounts for more than 98% of the total variance in the observed creep responses. The scores of the cells examined on the first PC axis increase with increasing cancer malignancy. They also correlate highly with the average moduli and viscosities extracted from viscoelastic models (-83% correlation with moduli and -85% correlation with viscosities). This suggests a direct link between the malignancy of cancer and the overall creep compliance level that is governed by cell viscoelastic properties. The implications of the results are discussed for the detection of non-tumorigenic and tumorigenic breast cells at different stages of cancer progression.