Mechanisms of Pluripotency In Vivo and In Vitro

Eszter Posfai *; Oliver H. Tam *; Janet Rossant *,†,1    
* Developmental and Stem Cell Biology, The Hospital for Sick Children Research Institute, Toronto, Ontario, Canada
† Department of Molecular Genetics, University of Toronto, Toronto, Ontario, Canada
1 Corresponding author: email address: janet.rossant@sickkids.ca

1 Introduction

At the onset of mouse development, fertilization of an oocyte by a sperm generates a 1-cell embryo, also known as a zygote. The zygote possesses totipotency, the ability to generate an entire organism, including its extraembryonic supporting tissues. As the zygote travels through the oviduct, it undergoes cleavage to form an 8-cell embryo around embryonic day 2.5 (E2.5) (Fig. 1.1). Individual 8-cell blastomeres are still totipotent (Kelly, 1977; Tarkowski, 1959; Tarkowski & Wróblewska, 1967). However, this period of totipotency does not persist, with the blastomeres gradually restricting their developmental fate. The first lineage restriction (and the subsequent loss of totipotency) begins with the compaction of the 8-cell embryo and ends in the formation of the inner cell mass (ICM) and the trophectoderm (TE) of the blastocyst at E3.5. The ICM will undergo another lineage segregation event prior to uterine implantation, whereby cells become restricted to either the primitive endoderm (PE) or epiblast (EPI) lineages (Arnold & Robertson, 2009; Schrode et al., 2013; Stephenson, Rossant, & Tam, 2012). These lineages become clearly distinguishable by the late blastocyst stage (~ E4.5), with the TE forming an outer layer of epithelium surrounding a fluid-filled cavity (blastocoel), the EPI forming a ball of cells on one side of the blastocyst, and the PE migrating to generate an epithelial layer of cells separating the EPI from the blastocoel cavity. From fertilization to the time of implantation, a period also known as preimplantation development, the single-cell totipotent zygote has segregated into three developmentally distinct lineages, each of which will contribute to discrete cell populations essential for embryonic growth and development.

Although there are clear molecular and morphological differences between the three cell populations (EPI, TE, PE) of the preimplantation embryo, their divergent developmental potential is more apparent after implantation (Nowotschin & Hadjantonakis, 2010). In mouse, the cluster of EPI cells in the blastocyst rapidly expands to form a cup-shaped single-layer epithelium known as the postimplantation EPI. These cells subsequently generate embryonic ectoderm, mesoderm, endoderm, and germ cells of the developing organism, and contribute extraembryonic mesoderm to structures such as the allantois, amnion, yolk sac, and chorion (Lawson, Meneses, & Pedersen, 1991). The ability of the EPI population to generate the three major embryonic germ layers and the germ line has led to its classification as the pluripotent population of the developing embryo. In contrast, the TE and PE populations contribute almost exclusively to the extraembryonic tissues, forming structures such as the placenta (Cross, 2005; Rossant & Cross, 2001) and the parietal yolk sacs (Arnold & Robertson, 2009; Rossant & Tam, 2009).

2 Phases of Lineage Restriction

A complex cellular process such as lineage restriction and determination is unlikely to be a single-step event. Molecular markers, some of which are lineage-determining transcription factors (TFs), allow us to separate lineage restriction into three phases: initiation, commitment, and maintenance. During the initiation phase, differences in cellular states arise, either “stochastically” or through subtle external cues, in the previously homogenous cell population (Johnston & Desplan, 2010). Despite these differences, these cells remain “plastic” and are able to interconvert between different cellular states depending on external stimuli. Early-response TFs are then expressed in cells that become biased toward a specific developmental fate and are often used to identify populations that progress toward lineage commitment. In the commitment phase, the cellular state is no longer “plastic,” and becomes refractory to the initial inducing signal. In turn, specific developmental programs are activated that become increasingly dependent on cell-intrinsic cues, as seen in germ cell differentiation (Magnúsdóttir, Gillich, Grabole, & Surani, 2012). This is typically achieved by the activity of the early-response TFs, which can act by (a) forming positive feedback loops to strengthen their expression and (b) activating downstream (late response) TFs that are hallmarks of the cellular state (Zernicka-Goetz, Morris, & Bruce, 2009). In the final maintenance phase, the cell faces the task of stably maintaining expression of its gene set and inhibiting reversion or conversion into inappropriate cell fates (Pietersen & van Lohuizen, 2008). In many cases (such as in stem or progenitor cells), a cell must also be responsive to cues that direct appropriate downstream differentiation. Therefore, many genes that are involved in the formation of differentiated cell types are likely to be poised for activation, awaiting the right signals to produce downstream lineages (Dillon, 2012).

Given the difficulties in studying lineage specification in vivo, there have been significant efforts to establish culture systems to study these events in vitro.

We are now able to derive and maintain several stem cell populations from the mouse embryo that recapitulate the three early developmental lineages in vitro. These include embryonic stem (ES) (Brook & Gardner, 1997; Evans & Kaufman, 1981; Martin, 1981) and epiblast stem cells (EpiSC) (Brons et al., 2007; Tesar et al., 2007), trophoblast stem (TS) cells (Tanaka, 1998), and extraembryonic endoderm cells (Kunath et al., 2005), which represent the EPI (ES and EpiSC), TE, and PE lineages, respectively. These cell lines serve as potentially useful models for studying the maintenance phase of their respective embryonic cell type as well as their downstream differentiation capabilities.

However, we must be cautious when interpreting observations of in vitro cellular states, as they have been artificially stabilized by culture conditions in the absence of various extracellular and positional cues that exist in an intact mouse embryo. Therefore, these cell lines, although invaluable, might be overly simplistic snapshots of in vivo states, and their adaptation to culture conditions might have resulted in de novo regulative properties.

In this review, we examine and...