(254a) Effects of Allelic Regulation On Nanog Expression Heterogeneity in Embryonic Stem Cell Populations

Nanog is a pluripotency transcription factor in embryonic stem cells (ESCs) and its expression pattern and function have been studied extensively in recent years. A distinct feature of ESC populations in vivo or in vitro is their heterogeneous expression of Nanog. A bimodal distribution of Nanog is observed with cells switching typically between states of NANOG^high and NANOG^low indicative of the robustness in Nanog expression heterogeneity. These observations have prompted the development of mathematical models featuring not only transcriptional fluctuations in Nanog but also feedback loops among Nanog and its partners, Oct4 and Sox2. However, Nanog was recently reported [1] to be allelically controlled in mouse ESCs (mESCs) adding another layer of regulation in the observed expression heterogeneity in addition to gene network activity. In this study, we developed a multiscale stochastic population balance equation (PBE) model for ESC ensembles to investigate the effects of allelic expression on Nanog heterogeneity.

A PBE system was constructed with each equation corresponding to a subpopulation of ESCs exhibiting a distinct allelic pattern of Nanog expression. The physiological state vector comprised cell size as indicator of cellular proliferation and the intracellular NANOG protein content. The model captures activities such as cell division and allelic regulation transpiring at longer time scales and fast occurring gene expression dynamics with transcriptional noise described by stochastic differential equations. Parameter calculation was based on experimental data carried out by our group [2] and others [1]. Numerical solutions to the PBE model were obtained via Monte Carlo simulations considering switching of the allelic pattern of Nanog expression as a Markov process.

Allelic regulation resulted in a trimodal distribution for Nanog. However, transcriptional noise caused the profile to become bimodal in agreement with findings from experiments by our group and others. The model also allows the calculation of the cell residence time in various modes of the distribution shedding light on the differential stability of the NANOG^low and NANOG^high states. Employing this framework, we showed that stem cell lines engineered with a reporter transgene knocked in an allelically regulated locus, exhibit nonmatching distributions of the endogenous and reporter proteins. This led us to investigate the performance of stem cell reporter systems with the transgene inserted in one or both alleles and with different or similar net production dynamics for the target and reporter proteins. The results show that the protein and reporter signatures are weakly correlated (Pearson product-moment correlation coefficient: 0.6) in typical stem cell lines with monoallelic insertion of the reporter gene. These findings point to pitfalls in the interpretation of the data from such reporter systems. Moreover, the PBE model presented here can be utilized to back-calculate the actual distribution of the native gene product from reporter signatures. Finally, the issue of whether stem cells with reduced Nanog expression can maintain their pluripotency under non-differentiating conditions is debatable. Aided by the PBE model, we analyzed data from stem cells with a single Nanog allele deletion. Our study shows that a subpopulation of these cells remains pluripotent while the deletion of a single Nanog allele does not simply reduce NANOG uniformly for all ESCs but modulates NANOG heterogeneity directly.

Multiscale models coupled to experimental studies will be essential in the design of efficient protocols for directing stem cell differentiation and for the reprogramming of cells [3]. Such models will also serve as a basis for the engineering and control of bioprocesses for the production of stem cell therapeutics.    

1. Miyanari Y, Torres-Padilla ME (2012) Control of ground-state pluripotency by allelic regulation of Nanog. Nature 483: 470-473.

2. Wu J, Tzanakakis ES (2012) Contribution of Stochastic Partitioning at Human Embryonic Stem Cell Division to NANOG Heterogeneity. PLoS One 7: e50715.

3. Wu J, Rostami MR, Tzanakakis ES (2013) Stem cell modeling: from gene networks to cell populations. Current Opinion in Chemical Engineering 2: 17-25.