Quantitative observation and modulation of cellular dynamics

Abstract

Stem cell research provides a useful tool for unraveling the molecular mechanisms and shedding light on cell transplantation therapies. Traditional qualitative experimental observation lacks the rich dynamic information, and therefore prevents obtaining the insight into cellular processes such as stochastic cell fate control and biological pattern formation. Quantitative approach helps promote research of these challenges. As passive measurements, the available commercial solutions have their own drawbacks and need to be improved to fulfill specialized requirements of the cutting edge studies. The multidisciplinary technology developed in this study provides multidimensional observation of cellular dynamic processes, automatic extraction of spatial-temporal quantitative information, and precisely active optical modulation of cell.

Here a novel RII autofocus system was developed to solve the focus drift problem by utilizing fiber collimated laser beam to reduce the beam size. The propagation properties of Gaussian beam indicate that high resolution could be achieved by minimizing focused Rayleigh length. The stability of live cell imaging system is improved by the RII system, which was demonstrated to have better focus resolution than the leading commercial solutions (1.3-5.3% of the objective DOF vs. 33-50%). When the home-built microscopy is incorporated with the RII system, it enables automatic and continuous observation of cellular dynamics to identify critical information, such as rare dedifferentiation and pattern formation mechanism. A novel fluorescent image processing algorithm was developed to automatically segment cells and extract dynamic information from vast observation data. The idea of mimicking human visualization was first applied in cell segmentation. A new criterion SILC was proposed for precise cell linking, and a novel evaluation of tracking results aiming at extracting effective information was firstly suggested in this study. This efficient algorithm was proved to have high average detection rate (94.5%) under different densities (100-2000 cells/〖mm〗^2), and high tracking result (at least 80%) of cells. Novel active light modulations were explored aiming at precisely modulating the exogenous gene expression of stem cells and quantitatively determining the motion properties of genetically programmed cells in extremely high density.

The development of biophotonics described in this thesis covers all the methodology holes in quantitative study of stem cell differentiation and pattern formation. This multidisciplinary approach is expected to help unravel critical dynamic properties of stem cells, and eventually contribute to make stem cell therapy a safe and effective option.