Single-Cell Analyses of ESCs Reveal Alternative Pluripotent Cell States and Molecular Mechanisms that Control Self-Renewal

Abstract

SummaryAnalyses of gene expression in single mouse embryonic stem cells (mESCs) cultured in serum and LIF revealed the presence of two distinct cell subpopulations with individual gene expression signatures. Comparisons with published data revealed that cells in the first subpopulation are phenotypically similar to cells isolated from the inner cell mass (ICM). In contrast, cells in the second subpopulation appear to be more mature. Pluripotency Gene Regulatory Network (PGRN) reconstruction based on single-cell data and published data suggested antagonistic roles for Oct4 and Nanog in the maintenance of pluripotency states. Integrated analyses of published genomic binding (ChIP) data strongly supported this observation. Certain target genes alternatively regulated by OCT4 and NANOG, such as Sall4 and Zscan10, feed back into the top hierarchical regulator Oct4. Analyses of such incoherent feedforward loops with feedback (iFFL-FB) suggest a dynamic model for the maintenance of mESC pluripotency and self-renewal.

Serum-Based Culture Conditions Provoke Gene Expression Variability in Mouse Embryonic Stem Cells as Revealed by Single-Cell Analysis

“…In agreement with recently published single cell analyses of mouse ESCs (Grün et al, 2014; Kolodziejczyk et al, 2015; Kumar et al, 2014; Papatsenko et al, 2015), we observed significant heterogeneity in gene expression in the serum cultured mouse ESCs. Using LLE analysis, we showed that heterogeneity does not appear to be stochastic, but rather follows a defined differentiation pathway towards PE-like cells.…”

“…More recent studies have also examined transcriptional networks and cell cycle regulators that contribute to transcriptional variation (Kolodziejczyk et al, 2015; Papatsenko et al, 2015). Epigenetic regulation, which may also contribute to overall variability, has not been adequately explored.…”

Inference of cell type specific regulatory networks on mammalian lineages

“…For each branch, a coarse skeleton regulatory network is inferred using GENIE3 [7], followed by detailed regulatory program learning using Gaussian Processes. While many of these approaches use single cell expression data alone, some approaches have integrated other sources of data, e.g., TF ChIP-seq [29] or TF knockdown assays [29, 30] to establish a network structure [29]. …”

Analysis of pluripotency transcription factor interactions reveals preferential binding of NANOG to SOX2 rather than NANOG or OCT4

“…ChIP-Seq studies show that Nanog often shares similar genomic binding sites with Sox2 (Papatsenko, Darr et al 2015). In addition, Sox2 and Oct4 synergistically bind the Sox-Oct cis co-14 motif (Ambrosetti, Basilico et al 1997, Mistri, Devasia et al 2015.…”