Automated laser ablation cell sorting and proteometrics of live cells in image cytometry

Live cell studies have been limited on the single cell level; thus many biological conclusions were drawn based on subjective observations of single micrographs, making statistical correlation of activity-function impossible. Therefore, an automated, high-throughput, quantitative image cytometer with autofocus was constructed to rapidly process thousands of cells in a monolayer.; A laser ablation system was incorporated into the cytometer to negatively sort out the unwanted cells by applying a focused, ultra-short laser pulse to each targeted cell. Due to the high power density (≈1010 W/cm2) that was present at the focal point, disruptive mechanical forces were generated and were responsible for the kill. A cell mixture model was studied in which faster dividing melanoma cells were fluorescently stained as model contaminant cells, whereas slower dividing background fibroblasts were unstained. Ablation pulses were applied in frame-by-frame increment batches to the cell culture on the microscope to purge the melanoma cells in the mixture. 99.6% confidence of total negative sorting effectiveness was found and confirmed by culturing the ablated areas for cancer cell re-growth.; With dye toxicity, phototoxicity, microscope optics, temperature, pH, humidity, and mechanical stability controls taken into consideration, the system was developed that allowed long-term cell studies while keeping extraneous environ mental perturbation to a minimum. Automatically, histone H2B-GFP fibroblasts were scanned for 72 hours and the histone content and inter-division times were measured to be 19.6 ± 6.0 hours and all the cell translocation tracks were traced.; The cytometer was modified to study p21 Rho GTPase protein activation localization in live mouse embryonic fibroblasts (MEF). The measurements were performed with a recently developed single-chain fluorescence resonance energy transfer (FRET) biosensor. The analyses were completely automated and involved K-mean clustering, cross-correlational alignment for image registration, ratiometric methods and invention of intuitive indices and plots to quantify the intracellular distribution of Rho GTPase activity. Activation patterns of Rho GTPase based on cell shape and polarization were classified to extract data correlating cell motility and direction of motion to Rho activity, indicating distinct localization patterns of Rho activation, prevalent in the directions of protrusion and retraction at the cell periphery.