Search for: All records

The metastasis of solid tumors hinges on cancer cells navigating through complex three-dimensional tissue environments, characterized by mechanical heterogeneity and biological diversity. This process is closely linked to the dynamic migration behavior exhibited by cancer cells, which dictates the invasiveness of tumors. In our study, we investigate tumor spheroids composed of breast cancer cells embedded in three-dimensional (3D) collagen matrices. Through a combination of quantitative experiments, artificial-intelligence-driven image processing, and mathematical modeling, we uncover rapid transitions in cell phenotypes and phenotype-dependent motility among disseminating cells originating from tumor spheroids. Persistent invasion leads to continuous remodeling of the extracellular matrix surrounding the spheroids, altering the landscape of migration phenotypes. Consequently, filopodial cells emerge as the predominant phenotype across diverse extracellular matrix conditions. Our findings unveil the complex mesoscale dynamics of invading tumor spheroids, shedding light on the complex interplay between migration phenotype plasticity, microenvironment remodeling, and cell motility within 3D extracellular matrices. Published by the American Physical Society2024 

Abstract Contemporary approaches to instance segmentation in cell science use 2D or 3D convolutional networks depending on the experiment and data structures. However, limitations in microscopy systems or efforts to prevent phototoxicity commonly require recording sub-optimally sampled data that greatly reduces the utility of such 3D data, especially in crowded sample space with significant axial overlap between objects. In such regimes, 2D segmentations are both more reliable for cell morphology and easier to annotate. In this work, we propose the projection enhancement network (PEN), a novel convolutional module which processes the sub-sampled 3D data and produces a 2D RGB semantic compression, and is trained in conjunction with an instance segmentation network of choice to produce 2D segmentations. Our approach combines augmentation to increase cell density using a low-density cell image dataset to train PEN, and curated datasets to evaluate PEN. We show that with PEN, the learned semantic representation in CellPose encodes depth and greatly improves segmentation performance in comparison to maximum intensity projection images as input, but does not similarly aid segmentation in region-based networks like Mask-RCNN. Finally, we dissect the segmentation strength against cell density of PEN with CellPose on disseminated cells from side-by-side spheroids. We present PEN as a data-driven solution to form compressed representations of 3D data that improve 2D segmentations from instance segmentation networks. 

Altered tissue mechanics is an important signature of invasive solid tumors. While the phenomena have been extensively studied by measuring the bulk rheology of the extracellular matrix (ECM) surrounding tumors, micromechanical remodeling at the cellular scale remains poorly understood. By combining holographic optical tweezers and confocal microscopy on in vitro tumor models, we show that the micromechanics of collagen ECM surrounding an invading tumor demonstrate directional anisotropy, spatial heterogeneity and significant variations in time as tumors invade. To test the cellular mechanisms of ECM micromechanical remodeling, we construct a simple computational model and verify its predictions with experiments. We find that collective force generation of a tumor stiffens the ECM and leads to anisotropic local mechanics such that the extension direction is more rigid than the compression direction. ECM degradation by cell-secreted matrix metalloproteinase softens the ECM, and active traction forces from individual disseminated cells re-stiffen the matrix. Together, these results identify plausible biophysical mechanisms responsible for the remodeled ECM micromechanics surrounding an invading tumor.