Text

theses

Two major challenges plague the robust design of stem cell-derived tissues for regenerative therapies: (1) the phenotypic and functional heterogeneity inherent in stem cell cultures, and (2) the dynamic, long-term nature of stem cell responses to microenvironmental cues. Several tools have emerged to precisely characterize how stem cells respond to various stimuli and scaffold properties; however, these tools are limiting because they are population-based and rely on the detection of markers expressed in fully differentiated cells. Thus, methods to characterize individual stem cells from a population is key for establishing cell-biomaterial relationships necessary to design scalable constructs for tissue engineering applications. This thesis dissertation focuses on the utility of single cell profiling techniques to identify the heterogeneity within cell cultures and characterize responses to controllable changes in diverse microenvironments. This method relies on the quantification of metrics derived from images of cytoskeletal and nuclear proteins that are sensitive to microenvironmental cues that influence cell state. This is achieved by pursuing two thesis-specific aims: (1) to utilize single cell biological imaging and machine learning techniques to identify cell subtypes in heterogeneous cultures, and (2) to use early morphological descriptors of intranuclear mechanotransductive proteins to predict long-term stem cell responses to biomaterials. In this study we report that single cell imaging-based profiling of cytoskeletal actin and nuclear mitotic apparatus (NuMA), a cell cycle regulating protein, can identify different cell phenotypes in heterogeneous stem cell cultures, progenitor cells derived from different tissue sections ex vivo, and stem cell responses to a diverse set of surface chemistries. We also show that the early (3 day) organization of interchromatin domains varies in human mesenchymal stem cells exposed to a variety of growth factor combinations and complex topographical microenvironments that induce long-term (> 7 day) divergent phenotypic outcomes. In summary, the results presented in this thesis dissertation show that single cell imaging-based profiling can be utilized to identify cell subtypes and predict microenvironment-induced differentiation fates at earlier times and with more resolution than current screening assays. This work can help lay the foundation for a new generation of single cell-based biomaterial screening tools and cellular phenotyping techniques.

ETD_6079

Ph.D.

Includes bibliographical references

by Sebastian L. Vega

doi:10.7282/T30V8FHV

ETD doctoral

The author owns the copyright to this work.