Integrative characterization on disease heterogeneity at single cell resolution

heterogeneity;single cell measurement;Chemical & Biomolecular Engineering

Chen, Wei-chiangMao, Hai-Quan ;

Johns Hopkins University

关键词: heterogeneity;    single cell measurement;    Chemical & Biomolecular Engineering;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/40652/CHEN-DISSERTATION-2014.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

Questions in cell biology are often answered with complexity of protein pathways. For years, cell biologists have been focusing on how molecular interactions affect cellular behaviors. However, when this concept is applied to disease, it sometimes deviates from what we expect. In general, the central dogma of biology starts from DNA, RNA, and then proteins, which bring about all the cellular functions. Extending from a single cell, it consists of more complex systems, like multiple cells and organism. By understanding the pathways of proteins in a single cell, there is certainly a gap at single cell level for understanding multiple cells and organism. In order to bridge the gap, I would like to emphasize the role of integrative characterization on single cell in this thesis.Integrative characterization means not only just measurement of protein expression but also simultaneous quantification of cellular/nuclear morphology. The system can provide better connection between molecular level and a cell level to better describe cellular behavior. First, to collect these information fast and robust, we developed a high throughput/high content image process assay which allows us to quantify cellular/nuclear morphology and protein expression and localization at single cell level. For each of single cells measured with this assay, there are around 600 parameters to describe their cellular and nuclear morphology and protein expression and localization. This assay is mainly used for most of the following work in this thesis. With the shape and size information extracted from the assay, we further applied principle component analysis and clustering analysis to demonstrate a phenotypic signature of metastasis in pancreatic cancers. In the exactly cell lines, geneticists show that there is not genetic signature for metastasis of pancreatic cancers. From our characterization, cells with higher metastatic potential are more homogeneous in their cellular/nuclear morphology. This signature can successfully predict cancer cells at different stages. The second part of this work, we targeted one essential biological process, epithelial to mesenchymal transition (EMT), with our assay. EMT plays a critical role in many different fields in biology, such as developmental biology, cancer biology, and stem cell etc. Conventionally EMT is characterized with molecular markers, such as E-cadherin, vimentin, and fibronectin. From western blot, the result can only suggest the protein expression of overall population. Our assay successfully demonstrates morphology has agreement with classical EMT markers in predicting epithelial and mesenchymal populations. Combined the result from long time laps live-cell microscopy, we discover that normal mouse mammary gland cells, a classical cell lines routinely used in EMT study, are dynamically switching their phenotypes from epithelial phenotypes to mesenchymal phenotypes back and forth without any chemical stimulation. It is totally different from the previous view on EMT. Cells were studied under chemical stimulation, usually TGF-β, and illustrated as a static analysis. The last part of this work discusses the role of an intrinsic factor, cell cycle, playing in single-cell analysis. Cell cycle is a serial steps needed to well-controlled to lead successful cell division. There are generally four different phases, G1, S, G2, and M phase, in a complete cycle. Cells have specific phenotypes when they stay in different cell cycle phases. The conventional assay to measure cell cycle is with flow cytometry, which needs to detach cells from culture environment. In order to understand the real-time condition, we applied our assay to simultaneously measure cell cycle and cellular/nuclear morphology at single cell level. The benefit of our assay is that there is no additional chemical applied to synchronize cell cycle, and the result can be directly linked to the cultural environment. Our finding also shows that cellular/nuclear morphology is highly dependent on cell cycle. Perturbation on one or the other can lead to a bias result. We propose a simple equation for quantification on the contribution from either cell cycle or intrinsic phenotypes. The result presented from this work aims to combine molecular expression and cellular and nuclear morphology to provide an overall view on cellular behaviors at single cell level. Through this integrative characterization, we hope to provide better understanding on cellular behaviors by combining molecular information and morphological information. Preliminary data suggests this type of measurement has potential to be applied on the study of complexity of epigenetic landscape at single cell level. Histone 3 acetylation is not directly correlated with nuclear morphology. Instead, coefficient variation of histone 3 acetylation correlated with coefficient variation of nuclear size. This result implies the possibility of using morphological information to predict epigenetic variation. Furthermore, preliminary data on nanoparticle uptake suggests that the mechanism of nanoparticle uptake is particle shape-dependent, and the incubation with nanoparticles has no effect on cell cycle progression. Our assay can provide more information to guide the design of nanoparticle and access the effect from the delivery of nanoparticles on single cells.