Genetic modification and mesodermal differentiation of human embryonic stem cells

Abstract: Background: Pluripotent human embryonic stem cells (hESCs) are a resource of great potential, available for research. Their capability to differentiate into any cell of the human body and an unlimited in vitro expansion offers the possibility to study the earliest human development and provides a source of cells for regenerative medicine. However, the ability to generate all cell types also makes it difficult to direct the process. Differentiation often appears to occur at random and the underlying molecular processes are poorly understood. Blood and bone cells are both applied in clinical cell therapies and originate from the same embryonic germ layer, the mesoderm. Although the differentiation of adult stem cells to specific cell types is extensively studied, the hESC model provides a possibility to follow the fate of more primitive cells. In order to steer the differentiation of hESCs, developmental transcription factors (TFs) can be induced through genetic modification. TFs control the expression of multiple genes and can potentially direct the development of blood and bone precursors. Specifically, the TFs HoxB4 and Osterix (Osx) have shown potential to influence the development of embryonic stem cells, albeit the mechanisms in early human mesoderm development are not well defined. Objective: To establish and examine hESC differentiation into blood and bone derivatives. Genetic modifications were introduced by lentiviral vectors to improve cell traceability and to evaluate the gain-of-function of HoxB4 and Osx. Results and conclusions: In the first study we achieved effective gene marking of hESCs using lentiviral vectors for transgene expression from a human Ubiquitin C promoter. Using cells marked with fluorescence, interactions between human feedercells and hESCs could be easily evaluated. As gene marking was stable and retained in any hESC derivatives, it allowed the studies of later developmental stages of cells overexpressing the transgene. In the last two studies, the lentiviral vector construction and the selection of modified cells permitted evaluation of different levels of transgene expression. In the second study, we established and characterized a basic bone differentiation model using several hESC lines. Two differentiation approaches were compared and both methods showed development into the osteogenic phenotype. Evaluation of the secreted extracellular matrix and deposited mineralized tissue showed that it resembled that found in bone. In the third study the effects of the transcription factor HoxB4 during early hematopoietic differentiation were evaluated by comparing; GFP, HoxB4low and HoxB4high over-expressing and unmodified hESCs, during embryoid body induced differentiation. HoxB4high cells showed an increased early hematopoiesis, while HoxB4low cells did not. Despite an upregluation of early hematopoietic markers in HoxB4high cells, markers for late blood maturation were absent. It was also determined that transgene expression increased during differentiation, which may have been one reason why the hemato-endothelial marker VE-cadherin was up-regulated instead of blood marker genes. In vivo teratoma analysis revealed no proper germ-layer formation from HoxB4high cells. In the fourth study the over-expression of the TFs Osx and HoxB4 was evaluated during osteogenic differentiation. Osx showed a similar dose-dependent behaviour as HoxB4. Low expression levels of Osx increased osteogenic differentiation but at high levels, surprisingly a more hematopoietic phenotype was induced. Furthermore, higher HoxB4 expression also induced osteogenic differentiation. The two last studies can be summarized to reveal a dose-effect of the TFs HoxB4 andOsx, whilst also presenting wider effects on multiple cell populations.