Mathematical modeling and simulation of stem cell differentiation and reprogramming by Waddington's epigenetic landscape

• Main Author: Guo, Jing
• Other Authors: Zheng Jie
• Format: Theses and Dissertations
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Computer science and engineering::Computer applications::Life and medical sciences
• Online Access: https://hdl.handle.net/10356/88079; http://hdl.handle.net/10220/45611
• Institution: Nanyang Technological University

Recently, the field of computational stem cell biology, a new subdiscipline of developmental biology, has received a remarkable level of attention from the researchers. It leverages computational algorithms and tools to study the mechanisms of regulating the behaviors of embryonic stem cells (ESCs). In particular, the invention of induced pluripotent stem cells (iPSCs) in 2006 raised a question of intense focus: “What are the underlying molecu- lar mechanisms by which iPSCs regain pluripotent potential?” Dynamical systems modeling has revealed the complexity of the regulation of cell fate determination and thus has proven to be a useful method for answering this question. Towards exploring the molecular basis of stem cell fate regulation, the transcriptional regulatory network of core pluripotency transcription fac- tors (TFs) was first modeled to uncover critical network structures that contribute to the cell fate transitions. Subsequently, we integrated the epi- genetic network and signaling pathways that support the maintenance of stem cells and the determination of lineage specifications with the TF-based genetic layer of regulation, thereby constructing a hybrid network model of multilayered regulations. However, most of the computational models are mainly limited to exploring the system dynamics by theoretical modeling. More recently, a trend has emerged in computational stem cell biology to in- corporate genome-wide sequencing data, especially single-cell profiling, with system-level modeling of stem cells. In this thesis, we concentrate on these two computational methodologies, i.e., the knowledge-based modeling and the data-driven modeling, to facil- itate the study of stem cell fate transitions using a well-known metaphor in development biology, namely Waddington’s epigenetic landscape. It de- scribes the outcome of the percolation of genetic lineages by linking the cell phenotypes with the genotypes, through the coordination of gene regulatory networks. Thus, the trajectories of cell state transition on the landscape surface reflect the progressive dynamics of stem cell fate determination. We developed a software tool named “NetLand” to facilitate the knowledge- based modeling and simulation study of cell state transitions driven by ki- netics of gene regulatory network in Waddington’s epigenetic landscape. It serves as a useful tool for understanding the dynamics of stem cells. For example, it can be used to study the barriers to the conversion of iPSCs. Although there have been many a plenty of protocols for the generation of iPSCs, the low conversion rate is still a bottleneck. To apply iPSCs in disease modeling and regenerative medicine, it is crucial to identify the key barriers that prevent the transition of cell types from a somatic state to an ESC-like state. NetLand identifies epigenetic blocks which have been shown to have an impact on the conversion rate of iPSCs, in the reprogramming process after analyzing a reconstructed hybrid network. The leverage of the next generation sequencing techniques using data- driven approaches dramatically benefits the study of molecular regulations in stem cells. Here, we proposed a data-driven method, named HopLand, to model the process of stem cell fate determination by using the continuous Hopfield Network (CHN) to map cells in a Waddington’s epigenetic land- scape. The combinatorial regulatory interactions among genes constitute a kinetic model which can be calibrated by learning from single-cell gene expression data. HopLand is applied to estimate the pseudotimes of indi- vidual cells from the single-cell data, demonstrating its potential to generate fundamental insights into cell fate regulation. The mathematical modeling methods proposed in this thesis, i.e., knowledge- based and data-driven approaches, yield insights into stem cell fate transi- tions. Subsequently, they have enhanced our capability to engineer cells, for example, increasing the conversion rate from somatic cell types to iPSCs. Many open challenges are still impeding cell fate engineering from broad clin- ical applications. We believe that the research described in this thesis will provide instrumental quantitative techniques to facilitate the understanding of cell dynamics and contribute to human health care.