II. PROPERTIES OF UNDIFFERENTIATED EMBRYONIC STEM CELLS
A. Mouse ES Cell Lines
Mouse ES (mES) cell lines were first established in the early 1980s (17, 98, 108, 221, 396). Initially, this required the isolation and cultivation of preimplantation embryos (blastocysts) on mouse embryonic fibroblasts (MEFs), followed by the expansion of primary ES cell outgrowths through careful enzymatic dissociation (trypsin/EDTA) and subculture regimes (see Ref. 301). The efficiency of ES cell derivation proved strain dependent, and inbred mice, like the 129 mouse strain, demonstrated the highest rates of success for the generation of ES cells (321). Once established, murine ES cell lines displayed an almost unlimited proliferation capacity in vitro (review in Ref. 333) and retained the ability to contribute to all cell lineages. In vitro, mES cells maintained a relatively normal and stable karyotype, even with continued passaging. ES cells were also characterized by a relatively short generation time of 
12–15 h with a short G1 cell cycle phase (310).
Because the generation of ES cell lines initially required a monolayer of inactivated MEFs, it was reasoned that fibroblasts provided some critical factor(s) either to promote self-renewal or to suppress differentiation. Two groups independently identified leukemia inhibitory factor [LIF (391); identical to the "differentiation inhibitory activity" DIA (334)] as the trophic factor responsible for this activity. LIF is a soluble glycoprotein of the interleukin (IL)-6 family of cytokines, which acts via a membrane-bound gp130 signaling complex to regulate a variety of cell functions through signal transduction and activation of transcription (STAT) signaling (review in Ref. 59). These cytokines, including IL-6, IL-11, oncostatin M (OSM), ciliary neurotrophic factor (CNTF) and cardiotrophin-1 (CT-1), all show similar properties with respect to the maintenance of pluripotency of mES cells (57, 250). The absence of IL-6 family members, the removal of MEFs, or the inactivation of STAT3, a downstream signaling molecule of the gp130 signaling complex, promote ES cells to spontaneously differentiate in vitro (39).
Studies on hematopoietic stem cell expansion had suggested that ligand-receptor complex thresholds of soluble cytokines could be maintained by high concentrations of soluble cytokines or by cytokine presentation on the cell surface. According to this model, when a relevant ligand-receptor interaction falls below a certain threshold, the probability of differentiation is increased; otherwise, self-renewal is favored. Examination of ES cells over a range of LIF concentrations demonstrated that LIF supplementation had little effect on growth rates, but it significantly altered the probability of cells undergoing self-renewal versus differentiation (414). To further address this question, a designer cytokine (a fusion protein of sIL6/sIL-6R linked to a flexible peptide chain) called Hyper-IL-6 (HIL-6) (118) together with LIF were employed to experimentally and computationally test their capacity to sustain ES cell self-renewal. Quantitative measurements of ES cell phenotypic markers, functional assays (EB formation), and transcription factor (Oct-3/4) expression over a range of LIF and HIL-6 concentrations demonstrated a superior ability of LIF to maintain ES cell pluripotentiality at higher concentrations (
500 pM). These results supported a ligand/receptor signaling threshold model of ES cell fate modulation that requires appropriate types and levels of cytokine stimulation to maintain self-renewal (375).
Identification of cell surface and molecular markers has proven critical to define the molecular basis of stem cell identity or "stemness." It is now well established that undifferentiated mES cells express specific cell surface antigens (SSEA-1; Ref. 336) and membrane-bound receptors (gp130; Refs. 57, 250) and possess enzyme activities for alkaline phosphatase (ALP; Ref. 396) and telomerase (review in Refs. 11, 277; see Table 1). ES cells also contain the epiblast/germ cell-restricted transcription factor Oct-3/4 (268, 318). In vivo, zygotic expression of this POU domain containing transcription factor is essential for the initial development of pluripotentiality in the ICM (247). In ES cells, continuous Oct-3/4 function at appropriate levels is necessary to maintain pluripotency. A less than twofold increase in expression causes differentiation into primitive endoderm and mesoderm, whereas loss of Oct-3/4 induces the formation of trophectoderm concomitant with a loss of pluripotency (251; see Fig. 4).
Recently, two groups identified the homeodomain protein Nanog as another key regulator of pluripotentiality (73, 233). In preimplantation embryos, its expression is restricted to and required in epiblast cells from which ES cells can be derived. The dosage of Nanog is a critical determinant of cytokine-independent colony formation, and forced expression of this protein confers constitutive self-renewal in ES cells without gp130 stimulation. Nanog may therefore act to restrict the differentiation-inducing potential of Oct-3/4.
Both Nanog and Oct-3/4 are essential to maintain ES cell identity, but STAT3, following LIF activation, plays an accessory role. LIF, when applied to serum-free ES cell cultures, is insufficient to maintain pluripotency or block (neural) differentiation. In combination with bone morphogenetic protein (BMP), LIF sustains self-renewal, multilineage differentiation, chimera colonization, and germ-line transmission properties. The critical contribution of BMP is to induce expression of Id ("inhibitor of differentiation") genes via the Smad pathway. Forced expression of Id genes liberates ES cells from BMP or serum dependence and allows self-renewal in LIF alone. Blockade of lineage-specific transcription factors by Id proteins enables the self-renewal response to LIF/STAT3 signaling (410). MEK/ERK signaling is also involved in ES cell self-renewal and differentiation. Inhibition of MEK/ERK by the MEK inhibitor PD098059 inhibits differentiation and maintains ES cell self-renewal in culture, and the expression of ERK and SHP-2 is thought to counteract the proliferative effects of STAT3 and promote differentiation (review in Refs. 58, 59). It however remains currently unclear how this pathway interacts with Nanog, Oct-3/4, and LIF signaling to regulate pluripotentiality (see Fig. 4).
Finally, a recent study has implicated Wnt-signaling pathways in the maintenance of ES cell pluripotency. Wnt pathway activation by a specific pharmacological inhibitor (BIO; 6-bromoindirubin-3'-oxime) of glycogen synthase kinase-3 (GSK-3) maintains the undifferentiated phenotype in both mouse and human ES cells and sustains expression of the pluripotent stage-specific transcription factors Oct-3/4 and Nanog (314). The reversibility of the BIO-mediated Wnt-activation in hES cells also suggests a practical application of GSK-3-specific inhibitors to regulate early steps of differentiation, which may prove valuable for the derivation of cells suitable for regenerative medicine.
The ES cell property of self-renewal therefore depends on a stoichiometric balance among various signaling molecules, and an imbalance in any one can cause ES cell identity to be lost. Other molecular markers potentially defining pluripotentiality include Rex-1 (163, 304), Sox2 (16), Genesis (353), GBX2 (75), UTF1 (254), Pem (112), and L17 (303). All of these have been shown to be expressed in the ICM of blastocysts and are downregulated upon differentiation; however, they are not exclusively expressed by pluripotent embryonic stem cells and can be found in other cell types in the soma. Their potential role in maintaining pluripotentiality or self-renewal remains to be determined.
B. Human ES Cell Lines
The techniques used to isolate and culture murine ES cells proved critical to the generation of human (h) ES cell lines from preimplantation embryos produced by in vitro fertilization (265, 293, 362) and after in vitro culture of blastocysts (349) (see Fig. 3). The resulting hES cells shared some fundamental characteristics of murine lines, such as Oct-3/4 expression, telomerase activity, and the formation of teratomas containing derivatives of all three primary germ layers in immunodeficient mice (295, 362). Similar to mES cells, hES cells maintained proliferative potential for prolonged periods of culture and retained a normal karyotype even in clonal derivatives (4). In contrast to mES cells, hES cells formed mainly cystic EBs (168) and displayed proteoglycans (TRA-1–60, TRA-1–81, GCTM-2) and different subtypes of stage-specific antigens (SSEA-3, SSEA-4), which were absent from mouse ES cell lines (Table 1).
Several potentially important differences exist between mouse and human ES cells. hES cells show a longer average population doubling time than mES cells [30–35 h vs. 12–15 h (4)]. With murine cells, it is possible to substitute the feeder layer of embryonic fibroblasts with recombinant LIF, which signals through the gp130 receptor subunit to activate STAT3 (see above and Fig. 4). In contrast, LIF is insufficient to inhibit the differentiation of hES cells (293, 362), which continue to be cultured routinely on feeder layers of MEFs or feeder cells from human tissues. The identity of the essential self-renewal signal(s) provided to ES cells by MEF feeder cells remains ill defined. Despite the recent finding of a functional activation of the LIF/STAT3 signaling pathways in hES cells, LIF is unable to maintain the pluripotent state of hES cells (91). The cultivation of hES cells on extracellular matrix proteins, such as Matrigel (a complex mixture of ECM proteins isolated from Engelbreth-Holm-Swarm tumor) and laminin with MEF-conditioned media (401), causes hES cells to express high levels of 
6- and 
1-integrins, which are involved in cell interactions with laminin (401). These results show that the application of extracellular matrix-associated factors can be employed to improve the culture and maintenance of pluripotent hES cells.
At the end of 2001, 70 hES lines had been established using feeder layers of mouse embryonic fibroblasts. This panel of cells, however, suffers from significant limitations, including possible murine retrovirus infections (from the feeder cells) that have rendered them inappropriate for therapeutic applications. As of December 2004, only 22 of the cell lines listed in the NIH register have been successfully propagated in vitro [see update of December 10, 2004 in (http://escr.nih.gov/)], and although 17 karyologically normal (euploid) hES cell lines derived from human blastocysts were recently generated that could be subcultured by enzymatic dissociation (87), these cells were also established on MEFs. Importantly, hES cell lines have now been cultivated both on human feeder cells to avoid xenogenic contamination (5, 295) and in the absence of feeder cells under serum-free conditions (205) as had been previously done for mES cells (411). These technological advances suggest that new hES cell lines free from potential retroviral infections will be prepared and that these cells, unlike most of those currently available, might be suitable for eventual therapeutic applications.
Although the principle techniques necessary to culture (up to 80 and more passages) and manipulate hES cells have been established [cell cloning (4), cryo-preservation (294), transfection (104), and gene targeting by homologous recombination (419)], other methods (single-cell dissociation and proliferation) are still not yet optimal. Because of the variabilities among human ES cell lines (growth characteristics, differentiation potential, and culturing techniques), it will be important to define a reliable set of molecular and cellular markers that characterize the undifferentiated pluripotent (stemness) or differentiated state of hES cells. Recent attempts to define molecular markers of undifferentiated cells, however, indicate a high degree of variability among four hES cell lines maintained in a feeder-free culture system (70) and examined after long-term culture (312).
Several properties and molecular markers of hES cells are listed in Tables 1 and 2, but it is evident that the present data do not allow an unambiguous molecular definition of pluripotent stem cell properties. The application of transcriptome profiling with proteomic analyses to ES cell lines may prove useful to define which lines and growth conditions are optimal for human ES cells in vitro (see sect. VI). This information will also be necessary to set standards for hES cell research (see Ref. 52) and to answer the question, how many ES cells are necessary for research and medical applications (for further information on properties of specific hES cell lines, their cultivation, and differentiation abilities, see Ref. 79).
C. ES Cells of Other Species
Pluripotent stem cell lines have been generated from livestock (review in Ref. 277) and model organisms, such as chicken (74, 258), hamster (97), rabbit (142, 320), and rat (51, 56, 166, 372); however, only mouse and chicken ES cells have proven capable of colonizing the germ line. Of special importance for human stem cell research is the establishment of ES cell lines from nonhuman primates [rhesus monkey (263, 363), common marmoset (Callithrix jacchus, Ref. 364), and cynomolgus monkey (Macaca fascicularis, Ref. 352)]. Monkey ES cells, characterized by typical markers of human ES and EC cells (Oct-4, SSEA-4, TRA-1–60, TRA-1–81), retain a normal karyotype and have a high differentiation capacity in vitro (187, 363). These properties may qualify these cell lines as alternative and substitute model systems for hES cell lines. Moreover, after in vivo parthenogenetic development of Macaca fascicularis eggs to blastocyst-stage embryos, a pluripotent monkey stem cell line (Cyno-1) has been established that showed all the properties of hES cells, such as high telomerase and ALP activity; expression of Oct-3/4, SSEA-4, TRA 1–60, and TRA 1–81; and the ability to differentiate into various cell lineages (377). Specifically parthenogenesis is the process whereby a single egg can develop without the presence of the male counterpart.
These results suggest that stem cells derived from parthenogenetically activated eggs may also provide a potential source for autologous therapy (in the female), thus bypassing the need for creating embryos. However, aberrant expression of imprinted genes, either increased expression of maternally imprinted genes or reduced expression of paternally imprinted genes, may limit the usefulness of parthenogenetic lines and their derivatives due to their abnormal or diminished proliferative capabilities (152).
Answers to all your Biology Questions
