(815f) Analyzing Population Dynamics of Embryonic Stem Cell Cycle Transition Through An Integrated Experimental and Modeling Approach

Embryonic stem cells (ESC) have the potential to be used in many therapeutic applications due to their ability to expand in an undifferentiated state while being able to differentiate to any cell type under specific conditions. One important characteristic vital to this behavior is the cell cycle. In somatic cells, the cell cycle is characterized by a lengthy G1 (growth) phase and a long doubling time. ESC are typically characterized by a shortened G1 phase, with an increased proportion of cells residing in the S phase (DNA replication). During differentiation, there is a gradual transition from the ESC cycle to mature cell cycle.   However, limited work has been done in the area of dynamics of cell cycle transition from undifferentiated to mature cells. Our objective is to better understand this transition by tracking the cell cycle dynamics of a differentiating population of hESC. In this project we are using an integrated experimental and computational approach to investigate cell cycle traits of pluripotent stem cells (PSC), both at the self-renewal stage and during direct differentiation towards pancreatic lineage.

An important component in the analysis of this system is the heterogeneity in cell population and cell cycle state of each individual cell. We developed a stochastic population modeling framework to predict single cell level information from population based information gathered from flow cytometry data. This model is a cellular automaton model, with the state of the cell being defined by the cell cycle phase which the cell is in (G1, S, or G2/M). The residence times for phases are described by probability distributions and their moments. Individual cells are tracked with time, and in this way the population dynamics are simulated. This model was used with experimentally derived cell synchronization data. Two different PSC lines, H1 hESC and Y1-16 iPS, derived from dermal fibroblasts, were synchronized in the G2 phase with Nocodazole and released back into the cell cycle. Information on the subsequent dynamics of cycle phase distribution was collected via DNA stain (propidium iodide) and flow cytometry. Having developed the model for self-renewal of hESC, we extended it to a differentiating population. Specifically, we directed differentiation towards pancreatic lineage, and experimentally analyzed the cell cycle dynamics during differentiation in addition to the synchronization behavior of the pancreatic progenitor phenotype. 

Our population model framework is able to accurately predict phase resident time distributions from experimentally synchronized hESC. PSC synchronization revealed low population heterogeneity in cell cycle transition rates as shown by limited desynchronization of the cycle phase distribution. This data was compared to the population model, and through this comparison probabilistic parameters and characteristics associated with ESC were determined. Only one combination of specific distributions and their associated moments, representing phase residence times, were able to recapitulate the experimental data, and these parameters led to very good agreement between the simulations and experimental dynamics. To validate the model, we performed synchronization/BrdU and CFSE to experimentally measure mean phase and doubling times, and show that these are in very good agreement with those predicted in the model. The validated population model was next extended to induced hESC differentiation towards pancreatic lineage. Differentiation results in changes in phase residence time distributions, with the most prominent effect being an increase in the G1 population.

Together, the population model and synchronization experiments are able to extract single cell information on cycle phase times and variability from simple population analyses. Furthermore, the model is able to predict cell cycle information and describe self-renewal and differentiation behavior to high accuracy. The cell cycle information obtained by applying this methodology can aid in better understanding ESC and help in deriving functional, non-tumorigenic mature cells from a progenitor population for use in therapeutic applications, specifically insulin producing β-cells for diabetes treatment.