Human embryonic stem cells assemble and fulfill their developmental destiny.

How do we study the endocrine cell differentiation of the placenta? The accepted course of action dictates that you use primary trophoblast cell cultures derived from human placentas. This is accomplished using methods to isolate cytotrophoblasts that possess the potential to spontaneously differentiate into endocrinologically active syncytial trophoblast within a prescribed period of time (1). Depending upon the nature of the scientific question, you may use cell lines established from human choriocarcinomas (e.g. Jeg3, Jar, or BeWo cells). If you are more adventurous, you could opt for trophoblast stem cell populations derived from the mouse or rat (2, 3). Each model system has its merits and limitations. Golos and Thomson and their colleagues (4, 5) have discovered intriguing new experimental approaches for studying the endocrine cell differentiation of human trophoblast cells. These investigators have developed strategies for differentiating human embryonic stem (ES) cells into trophoblast cells capable of synthesizing and secreting placental hormones. Under routine culture conditions, human ES cells exhibit a modest capacity to spontaneously differentiate along the trophoblast lineage (6). Two methods have now been established to promote trophoblast cell differentiation from human ES cells: 1) treatment with bone morphogenetic protein (BMP) family members (4); and 2) aggregation and long-term culture on or within Matrigel, as described in this issue of Endocrinology (5). These two experimental manipulations lead to the in vitro formation of syncytial trophoblast, and their corresponding production of chorionic gonadotropin (CG), progesterone, and estradiol-17 , mimicking endocrine features of human trophoblast cells developing in situ. To place these exciting new investigative tools into the appropriate context, it is necessary that we delve into the biology of ES cells. First of all, it is important to appreciate that an ES cell is an experimental outcome of creating culture conditions for ex vivo propagation of stem cells isolated from blastocysts. Although blastocyst-derived cell populations from the human and mouse have been called ES cells they are not equivalent. Human and mouse ES cells differ in their derivation, the mechanisms they use for self-renewal, and their developmental potential, especially regarding their propensity to form cells of the trophoblast lineage (Fig. 1; Refs. 7 and 8). Human ES cells can differentiate into structures represented by each of the embryonic germ layers, as well as extraembryonic endoderm and trophoblast (6). In contrast, two types of stem cells have been isolated from the mouse blastocyst: 1) ES cells and 2) trophoblast stem (TS) cells. Mouse ES cells most closely correspond to the postimplantation epiblast, are pluripotent and capable of reconstituting all cell types of the body (7–9), whereas the developmental potential of TS cells is restricted to the trophoblast lineage (3, 7). Culture conditions for establishing human TS cells have not been reported. Unlike human ES cells, mouse ES cells do not routinely exhibit a capacity for trophoblast cell differentiation. This species difference is not an entirely new concept. Hints of a species difference between human and mouse stem cells first emerged during investigations with embryonal carcinoma cells over two decades ago (10, 11). Embryonal carcinoma cells are stem cells with an embryonic phenotype and are derived from teratocarcinomas. Human embryonal carcinoma cells can differentiate into trophoblast cells, a characteristic not usually shared with mouse embryonal carcinoma cells. A few exceptions exist to this view of the developmental potential of mouse ES cells. Oct4 is a POU family transcription factor associated with maintenance of stem cell status. Forced down-regulation of Oct4 in mouse ES cells yields spontaneous formation of trophoblast cells (12, 13). Such observations are consistent with the findings that Oct4 actively suppresses CG subunit gene expression (14, 15) and support the notion that Oct4 is a pivotal inhibitor of trophoblast development (13, 16). Unexpectedly, recent studies have demonstrated that mouse ES cells in long-term culture ( 43 d) can form blastocyst-like structures and express trophoblast-specific genes (17). Long-term mouse ES cell cultures may allow for reprogramming of the developmental potential of the stem cells. There is also a suggestion that some mouse ES cell lines may harbor a minute population of so-called “pre-TS cells” (18). The pre-TS cell is hypothesized to be a precursor of TS cells and ultimately the trophoblast lineage. Deficiency in the nuclear enzyme poly(ADP-ribose) polymerase-1 results in spontaneous differentiation of a small proportion of mouse ES cells into trophoblast derivatives, possibly through promoting survival and differentiation of the pre-TS cells (18). Thus, under a few special circumstances it appears that mouse ES cell populations can serve as progenitors for trophoblast cells. In contrast to mouse ES cells, human ES cells can be efficiently directed to differentiate into trophoblast cells. Exposure of human ES cells to BMP family members results in the morphologic and endocrine differentiation toward a trophoblast cell phenotype (4). The nature of the culture conditions is critical for BMP responsiveness. Monolayer cultures of human ES cells plated on Matrigel and exposed to mouse embryonic fibroblast conditioned medium (MEF-CM) and basic fibroblast growth factor (FGF2) respond to BMPs by differentiating into trophoblast within a few days. Alternatively, cells dissociated from human ES cell aggregates called embryoid bodies (see below) and cultured on fibronectin respond differently to BMPs; instead genes representative of Abbreviations: BMP, Bone morphogenetic protein; CG, chorionic gonadotropin; ES, embryonic stem; FGF2, basic fibroblast growth factor; MEF-CM, mouse embryonic fibroblast conditioned medium; TS, trophoblast stem.