Development of microscope-based automated systems for high-throughput imaging of cell dynamics

Abstract

Many biological processes in multi-cellular organisms are regulated by various dynamic cellular processes. However, classical tissue culture system can only provide static environments where many dynamic events occur randomly, thus mechanistic study of cellular dynamics has been limited. In this thesis, to overcome this limitation, we developed microcscope-based automated systems to enable high-throughput imaging and analysis of cell dynamics. First, we developed new experimental platforms where we can precisely position different types of cells and control their movement by illuminating light to predefined areas. Furthermore, we automated data acquisition by controlling individual modules of wide-field microscopes and attempted to automate data analysis by developing image processing algorithms. Finally, we applied the developed systems for the study of cell spreading and small-sized multi-cellular cluster dynamics. To control cell dynamics by light, we synthesized and characterized a cell-friendly photoresist (poly(2,2-dimethoxy nitrobenzyl methacrylate-r-methyl methacrylate-r-poly(ethylene glycol) methacrylate) (PDMP). Upon UV exposure, PDMP thin films become soluble in near-neutral aqueous buffers such as PBS and tissue culture media with minimal cytotoxicity. To create micro-patterns on PDMP thin film, microscope projection photolithography (MPP) technique was adopted. By performing series of MPP on PDMP thin films and sequentially depositing cells on UV-irradiated regions, we could successfully create single cell arrays of adhering cells. By controlling a motorized stage and excitation/emission filter wheels, dynamics of cells in single cell arrays were individually controlled and high-throughput imaging of cells were performed. Using this high-throughput imaging system, we first studied dynamics of cell spreading. To visualize substrate-contacting areas of cells, interference reflection microscopy (IRM) was performed. Since images acquired by IRM exhibited excellent contrast, automated data processing to extract spreading areas of cells was possible. Using this method, dynamics of about 80 spreading cells was observed in a single experiment and automatically analyzed. Furthermore, spreading of cells treated with six pharmacological inhibitors targeting cytoskeletons and their regulators was quantitatively analyzed. Next, we fabricated small-sized multi-cellular clusters with precisely controlled sizes, compositions, and geometries, and systematically investigated their dynamics. We found that the behavior of small-sized multicellular clusters was not sensitive to initial configurations, but was determined by dynamic force balances among the cells. Initially, the multicellular clusters exhibited a rounded morphology and minimal translocation, probably due to contractility at the periphery of the clusters. For 2-cell and 4-cell clusters, single leaders emerged over time and entire groups aligned and comigrated as single supercells. Such coherent behavior did not occur in 8-cell clusters, indicating a critical group size led by a single leader may exist.