Several reports have documented the generation of cells with hepatocyte characteristics in ES cell differentiation cultures. Hamazaki et al. (2001
) developed a multistep protocol that included the addition of specific growth factors at various stages of differentiation to promote the growth and differentiation of hepatocyte cells within the cultures. Genes indicative of hepatocyte development and maturation were expressed in these cultures with a kinetic pattern similar to that found in vivo. Hepatocyte-like cells generated with this protocol were subsequently shown to contain albumin protein and to produce urea (Chinzei et al. 2002). Cells from these cultures were also transplanted into recipient mice pretreated with 2-acetylaminofluorene to prevent proliferation of host hepatocytes. Four weeks following transplantation, low numbers of albumin-producing donor cells were detected in the livers of the recipient animals, suggesting that these cells might be able to function in vivo (Chinzei et al. 2002).
To be able to specifically monitor hepatocyte development and isolate these cells from the differentiation cultures, several studies have incorporated approaches to distinguish definitive endoderm and early liver populations from visceral endoderm. Jones et al. (2002) used an ES cell line carrying a lacZ gene trap vector inserted into a gene known as Gtar. Gtar is expressed in definitive endoderm committed to a hepatic fate and the developing liver but is not found in visceral endoderm. Following serum induction, LacZ-positive cells that coexpressed 
-fetoprotein (afp) and albumin (alb) were detected, indicating development of the hepatocyte lineage in the culture. In a recent study, Asahina et al. (2004) identified cytochrome P450 7A1 (Cyp7a1) as a liver-specific gene that is not expressed in visceral endoderm. Cyp7a1 expression was detected at low levels in serum-stimulated EBs at 2 wk of differentiation. As the Cyp7a1-positive cells were not isolated or characterized in this study, their developmental potential remains to be determined. Yamada et al. (2002a) used specific uptake of the organic anion indocyanine green (ICG) as a marker of hepatocyte development in ES differentiation cultures. ICG-positive cells isolated from the cultures expressed hepatocyte markers, displayed ultrastructural characteristics similar to those of hepatocytes and gave rise to low numbers of albumin-positive cells following transplantation into recipient animals. Taken together, the findings from these various studies strongly suggest that cells with hepatocyte characteristics are generated in mouse ES cell differentiation cultures.
Cells with hepatocyte characteristics have also been identified in hES cell cultures (Rambhatla et al. 2003). To generate these cells, the cultures were treated with sodium butyrate, a procedure that killed significant numbers of the differentiating population. The cells that survived this treatment gave rise to a population that displayed many features of hepatocytes. Although the approach is promising and results in the development of hepatocyte-like cells, the physiological relevance of the sodium butyrate treatment is unclear.
The findings from these different studies indicate that cells with characteristics of pancreatic 
-cells and hepatocytes can be generated in ES cell differentiation cultures. A problem with many of the current approaches is the very low frequency of differentiated cells identified and the cellular heterogeneity within the cultures. Improvements in generating endoderm derivatives will require the development of more efficient protocols for endoderm induction and for pancreatic and hepatocyte specification combined with technologies for selecting the appropriate cell types from the differentiation cultures. Only when large numbers of highly enriched progenitors are accessible can methods for their maturation be defined and their functional capacity be rigorously tested in animal models of diabetes and liver failure.
We have recently investigated the potential of ES cells to differentiate to endoderm derivatives and developed two different protocols that promote the generation of these cell types (Kubo et al. 2004). The first is a restricted exposure of the EBs to serum followed by a period of serum-free culture, and the second is induction with activin A (activin) in the absence of serum. Endoderm development was quantified based on the proportion of cells that expressed Foxa2, a transcription factor found in the earliest stages of definitive endoderm development (Monaghan et al. 1993; Sasaki and Hogan 1993). All of the Foxa2+ cells that developed in these cultures also expressed the primitive streak marker brachyury, a gene that is not expressed in visceral endoderm. This observation strongly suggests that the Foxa2+ cells represented definitive endoderm. Based on the number of Foxa2+ cells, the activin protocol was found to be the most efficient as >50% of the total population in these cultures expressed this protein. Tissue specification was detected in the activin-induced cultures as demonstrated by the presence of cells that expressed Afp and Alb, indicative of hepatocyte differentiation, Pdx1 for early pancreas specification, and surfactant protein C (Sp-c) for early lung development (Wert et al. 1993). When transplanted under the kidney capsule of recipient mice, the activin-induced cells generated lung-like structures that expressed Sp-c and gut structures that expressed intestinal fatty acid-binding protein (IFABP). These findings demonstrate that it is possible to efficiently generate endoderm in ES differentiation cultures with the use of specific inducers in the absence of serum.
Ectoderm derivatives
Ectoderm differentiation of mouse ES cells is well established, as numerous studies have documented and characterized neuroectoderm commitment and neural differentiation. Given extensive efforts in this field over the past decade, several different protocols have evolved to promote neuroectoderm differentiation. The various approaches include (1) treatment of serum-stimulated EBs with retinoic acid (Bain et al. 1995), (2) sequential culture of EBs in serum followed by serum-free medium (Okabe et al. 1996), (3) differentiation of ES cells as a monolayer in serum-free medium (Tropepe et al. 2001; Ying et al. 2003b), and (4) differentiation of ES cells directly on stromal cells in the absence of serum (Kawasaki et al. 2000; Barberi et al. 2003). As with the mesoderm and endoderm lineages, development of the ectoderm lineages in the ES differentiation cultures appears to recapitulate their development in the early embryo (Barberi et al. 2003). Commitment to neuroectoderm appears to be rapid and efficient as the majority of the cells that develop in these cultures display characteristics of neural cells (Okabe et al. 1996; Kawasaki et al. 2000; Barberi et al. 2003). By targeting the -geo selection marker gene to Sox2, a gene expressed in neuroepithelium, Li et al. (1998) were able to generate highly enriched neural populations (>90% of the population) using the retinoic acid-induction protocol followed by selection with G418.
Each of the three major neural cell types of the central nervous system—neurons, astrocytes, and oligodendrocytes—can be generated, and relatively pure populations of each can be isolated when cultured under appropriate conditions (Okabe et al. 1996; Barberi et al. 2003). In addition to the generation of these different neural populations, conditions have been established for the development of different subtypes of neurons. The protocols for differentiation to specific types of neurons have included the sequential combination of regulators that are known to play a role in the establishment of these lineages in the early embryo. For instance, midbrain dopaminergic neurons have been generated in the EB system by overexpression in the cells of the transcription factor nuclear-receptor-related factor1 (Nurr1), and the addition to the cultures of SHH and FGF8 (Kim et al. 2002). Nurr1, SHH, and FGF8 are required for the development of this class of neurons in the early embryo (Ye et al. 1998; Simon et al. 2003). More recent studies have demonstrated the development of cholinergic, serotonergic, and GABAergic neurons in addition to dopaminergic neurons, when differentiated on MS5 stromal cells in the presence of different combinations of cytokines (Barberi et al. 2003). Using the coculture approach together with the appropriate signaling molecules and selection steps, Wichterle et al. (2002) successfully generated cells that display many of the characteristics of motor neurons. As with many other populations discussed in this review, the generation of these cells from ES cells recapitulated the pathway of motor neuron development in vivo.
The ES cell model is being used to investigate the earliest stages of neural development. When cultured at low density in serum-free medium in the presence of LIF, ES cells generate a population that has been called primitive neural stem cells (Tropepe et al. 2001). These cells have been characterized by their ability to generate neurosphere-like colonies composed of cells that express the neural precursor cell marker, nestin (Lendahl et al. 1990). When cultured on a matrigel substrate in the presence of low amounts of serum, cells within these colonies generated neurons, astrocytes, and oligodendrocytes. Mesoderm and definitive endoderm derivatives were not detected in these cultures. In contrast to the restricted developmental pattern observed in culture, the cells from these colonies were able to contribute to most tissues of the embryo following morula reaggregation and transplantation in vivo. These observations would position the primitive neural stem cell at a developmental stage between the ES cell and a neural-restricted progenitor.
The ability to generate different types of neurons from ES cells has dramatically raised the interest in repair of nervous system disorders by cell replacement therapy. Early transplantation studies demonstrated the feasibility of engrafting ES-cell-derived neural cells into recipient animals. Brustle et al. (1997) transplanted such cells into ventricles of fetal rats and demonstrated the incorporation of donor-derived neurons, astrocytes, and oligodendrocytes into the brains of the recipient animals, several weeks later. As a demonstration of cell-based repair, this group showed that ES-cell-derived oligodendrocytes could form myelin sheaths around host neurons when transplanted into a myelin-deficient rat model of multiple sclerosis (Brustle et al. 1999). Other studies demonstrated that ES-cell-derived neural cells could lead to partial functional recovery of spinal cord injury in rats, when injected directly into the spinal cord near the injury site (McDonald et al. 1999). Although the basis for recovery in this study was not determined, remyelination of the damaged nerves may have played some role, as a later study demonstrated that ES-cell-derived neural cells transplanted into spinal cords of rats with chemically induced demyelination preferentially differentiate into oligoendrocytes and myelinated host axons (Liu et al. 2000).
One of the greatest expectations of clinical utility for ES-cell-based therapy for neurodegenerative disease is the replacement of dopamine neurons in Parkinson's patients. Using a rat model for Parkinson's disease, Kim et al. (2002) demonstrated that ES-cell-derived dopamine neurons survived, developed functional synapses, and displayed electrophysiolgic properties characteristic of midbrain neurons following transplantation into these animals. In addition, the animals showed some recovery, suggesting that the transplanted cells were functional. The animals that received cells overexpressing Nurr1 showed the greatest improvement in this study. Populations not overexpressing exogenous Nurr1 generated with the stromal cell coculture protocol have also been transplanted into Parkinsonian mice (Barberi et al. 2003). Donor dopamine neurons were detected 2 mo following transplantation, and grafted animals showed alleviation of the behavioral deficits displayed by the control animals. Taken together, these studies establish that it is possible to generate specific subsets of neural cells from mouse ES cells and that these cells can be transplanted into models of human diseases. Additional studies will be required to determine the extent to which these cells can function over extended periods of time.
In addition to the neural lineages, ES cells can also give rise to epithelial cells that will undergo differentiation to populations that express markers of keratinocytes (Bagutti et al. 1996). The temporal sequence of expression of different keratins associated with development of the lineage in culture is very similar to patterns found in the mouse embryo. When maintained for 3 wk in organ cultures, the ES-cell-derived keratinocytes were able to form structures that resemble embryonic mouse skin, indicating that they possess some capacity to generate a functional tissue (Coraux et al. 2003). Differentiation of the epidermal cells appears to be controlled, in part, by BMP4 (Kawasaki et al. 2000). When added to cultures undergoing neuroectoderm differentiation, BMP4 promoted keratinocyte development and inhibited neural differentiation.
Answers to all your Biology Questions
