Genomic approaches to deconstruct pluripotency

Abstract

Embryonic stem cells (ESCs) first derived from the inner cell mass of blastocyst-stage embryos have the unique capacity of indefinite self-renewal and potential to differentiate into all somatic cell types. Similar developmental potency can be achieved by reprogramming differentiated somatic cells into induced pluripotent stem cells (iPSCs). Both types of pluripotent stem cells provide great potential for fundamental studies of tissue differentiation, and hold promise for disease modeling, drug development, and regenerative medicine. Although much has been learned about the molecular mechanisms that underlie pluripotency in such cells, our understanding remains incomplete. A comprehensive understanding of ESCs and iPSCs requires the deconstruction of complex transcription regulatory networks, epigenetic mechanisms, and biochemical interactions critical for the maintenance of self-renewal and pluripotency. In this review, we will discuss recent advances gleaned from application of global "omics" techniques to dissect the molecular mechanisms that define the pluripotent state.

Figure 1

Transcriptional circuitry that maintains pluripotency. (a) The key transcription factors of pluripotency form positive reciprocal and autoregulatory loops that maintain the expression of Oct4, Sox2, and Nanog. The key factors also synergistically co-occupy numerous downstream target genes that promote self-renewal and maintain pluripotency, while repressing developmentally regulated genes that drive differentiation. Transcription factors are represented by ovals, and the genes are represented by rectangles. (b) OCT4, SOX2, and NANOG co-occupy the multiple transcription-factor-binding loci (MTL) enhancer and positively regulate numerous noncoding RNAs in human induced pluripotent stem cells (iPSCs). Large intergenic noncoding RNA–regulator of reprogramming (LincRNA-RoR), a downstream effector of the core network, is important for establishing pluripotency during iPSC reprogramming. (c) An Oct4-centered network in ESCs. A schematic network, constructed based on Reference , consists of Oct4-interacting proteins and interacting partners of Oct4-associated proteins. Complexes consisting of several protein subunits are indicated by large yellow circles. The Oct4 interactome was further wired to the transcription regulatory network through integration of data sets from microarray profiling and transcription factor binding. Rectangular nodes represent genes that are bound by Oct4 as reported by previous ChIP-on-chip or ChIP-seq studies (18, 52). Red indicates functional regulation, as the expressions of respective genes were repressed with reduced levels of Oct4 in ZHBTc4 ESCs (104). Thick blue lines connect Oct4 with transcription factors that synergistically co-occupy downstream target genes with Oct4.

Figure 2

Genomic-wide mapping of protein-DNA interactions, histone modifications, and DNA methylation. Transcriptional networks and characteristics of the epigenome such as histone modifications and DNA methylation can be uncovered by highly condensed microarray chips or by next-generation sequencing technologies, respectively.

Figure 3

Crosstalk between histone and DNA modifications. (a) DNA methylation can direct either acetylation or H3K9 methylation. DNA methyltransferases (DNMTs) have been found to associate with both histone deacetylases (HDACs, left) as well as H3K9 methyltransferases (right—e.g., G9a). DNA methyl binding proteins (e.g., MECP2) associate with both HDACs and H3K9 methyltransferases (e.g., ESET). (b) H3K9 methyltransferases can direct DNA methylation. The SUV39H1/2 and G9a histone methyltransferases (HMTs), when complexed with HP1 (adapter), can recruit DNMTs (left). ESET can complex with DNMTs directly (right). (c) H3K27 methyltransferase EZH2 can direct DNA methylation. DNMTs have been found to bind directly to EZH2.