br Experimental Procedures br Author Contributions br
Introduction The defining features of pluripotent stem gsk-3 (PSCs), whether they originate from germ cell tumors, from the embryo, or through cellular reprogramming, are their abilities to undergo self-renewal and to give rise to all of the tissues of the body. However, this straightforward operational definition of pluripotency has been complicated in recent years by the revelation that there are a number of distinct cellular states that display these features. In the mouse, the species in which our understanding of PSCs is most advanced (Nichols and Smith, 2012; Tesar et al., 2007), there are two widely recognized states of pluripotency, referred to as naive and primed states, corresponding to distinct stages of peri-implantation embryonic development. Strong pharmacological suppression of the primary signaling pathways that drive differentiation enables the maintenance of mouse embryonic stem cells (ESCs) from the preimplantation epiblast in a naive state of pluripotency, defined as a fully unrestricted state that possesses the flexibility to give rise to all embryonic lineages and to form germline chimeras (Ying et al., 2008). PSCs isolated from a later stage of development, the postimplantation epiblast, are known as epiblast stem cells (Brons et al., 2007; Tesar et al., 2007). These cells lack the ability to form chimeras when introduced into preimplantation embryos but will give rise to teratomas when injected into host animals and can colonize all tissues including the germline when assayed in postimplantation embryo cultures in vitro (Huang et al., 2012). Besides the disparity in developmental potential in vivo, there are other significant differences between these two types of PSCs, both in terms of gene expression and their requirements for stem cell maintenance. Importantly, epiblast stem cells display more marked expression of genes associated with early germ layer formation (Tesar et al., 2007). The question of what development state primate ESCs equate to has never been clearly resolved. Early work on cell lines from human germ cell tumors, confirmed by studies on monkey and human ESCs, showed clearly that primate PSCs differ in phenotype from mouse teratocarcinoma or mouse ESCs (Pera et al., 2000). By contrast, mouse epiblast stem cells resemble human ESCs in many respects. However, there are also some significant differences between these two cell types. Gafni et al. (2013) recently reported cell-culture conditions that support maintenance of human PSCs in a naive-like state, with high levels of pluripotency-associated gene expression, minimal expression of lineage-specific genes, and a high capacity for self-renewal. Chan et al. (2013) also described conditions that support maintenance of naive human PSCs, which showed strong coexpression of GATA6 and NANOG, similar to epiblast cells. The cell types described by these two groups were similar to mouse naive PSCs but were different in some aspects, in particular, in their requirement for nodal/activin and FGF signaling for stem cell maintenance. Efforts to understand the states of pluripotency in different species are complicated by heterogeneity in ESC and epiblast stem cell lines, and by the existence of subpopulations of cells in both mouse and human ESC cultures that display lineage priming, or the coexpression of pluripotency and lineage-specific genes (Enver et al., 2009; Martinez Arias and Brickman, 2011; Nichols and Smith, 2009). Though the occurrence of heterogeneity in ESC populations in vitro and in the embryo in vivo is now widely accepted, recent results on mouse ESCs challenge the notion that it is an inherent feature of the pluripotent state (Marks et al., 2012). Marks et al. (2012) have shown that compared to cells maintained in serum-supplemented medium, in mouse ESC cultures strictly maintained in a naive state of pluripotency, heterogeneity in expression of key pluripotency genes was vastly reduced, coexpression of pluripotency and lineage-specific genes was strongly suppressed, and the bivalent chromatin marks seen in cells grown under conventional conditions, thought to reflect a type of molecular priming for differentiation, are reduced. Thus, recent debate has focused on whether heterogeneity is inherent to PSCs, or whether it is simply a function of the microenvironment of the stem cell under particular conditions of growth in vitro (MacArthur and Lemischka, 2013; Smith, 2013).