SEARCH

SEARCH BY CITATION

Keywords:

  • Activin;
  • ES cell;
  • organogenesis;
  • organizer;
  • Xenopus

Abstract

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

Studies performed over the last century have clarified the mechanisms of organ and tissue formation. Mesoderm formation is one of the most important events in early body pattern determination during embryogenesis. In 1988, we found that activin A has mesoderm-inducing activity. As activin A could induce dorsal mesoderm formation, unlike fibroblast growth factor and bone morphogenetic protein, this factor was thought to be the molecular entity of the Spemann-Mangold organizer. Subsequently, the mechanisms of early embryogenesis have been clarified using molecular biological techniques, resulting in the identification of many genes that are involved in organ and tissue development. This finding that activin A could induce dorsal mesoderm formation spurred research into the application of agents that induce organs and tissues in vitro. In this regard, we have shown that many organ types can be induced by activin A in vitro. Moreover, we have found that other types of organs can be induced by changing the conditions of treatment. To date, more than 20 different types of tissues and organs have been successfully induced from Xenopus undifferentiated cells in vitro. In recent years, we have applied these protocols to mouse embryonic stem cells, and we have successfully induced several tissues, such as the pancreas and cardiomyocytes. We are also investigating how the pluripotency of undifferentiated stem cells is regulated. In this review, we summarize the current knowledge regarding activin as a mesoderm-inducing factor and its application for the induction of tissues and organs from undifferentiated cells. Moreover, we provide some examples of in vitro tissue differentiation from mouse embryonic stem cells, which may prove useful in regenerative medicine.


Activin as a mesoderm-inducing factor

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

In vertebrates, early embryogenesis is a dynamic event that involves many complicated processes. Numerous factors play essential roles in these processes, thereby ensuring that the minute and precise formation of tissues and organ is achieved. Mesoderm formation is one of the most important biological events for embryonic patterning. In the 1930s, Hans Spemann and Hilde Mangold proposed the “organizer region” based on elegant experiments they performed in amphibian embryos, in which transplantation of the dorsal lip into the ventral blastocoele region induced secondary axis formation (Spemann & Mangold 1924). From this finding, the “double potential hypothesis” (Yamada 1938, 1958) and the “double gradient theory” (Toivonen & Saxen 1955) for embryonic pattern formation have been developed. These models propose that two presumptive factors form a concentration gradient in the embryo and determine body axes. Thus, it has been suggested that fundamental embryonic patterning with mesoderm formation is determined by only a few factors, and many researchers have expected to identify a highly restricted set of mesoderm-inducing factors. However, since the isolation of these factors has proven difficult, some investigators have come to the conclusion that this putative factor is not unique in inducing mesoderm formation or that it does not exist.

Nevertheless, during the 1980s, several factors that possessed mesoderm-inducing activities were identified. For example, basic fibroblast growth factor (bFGF) shows mesoderm-inducing activity (Slack et al. 1987), and in 1988, transforming growth factor (TGF)-β2 was also shown to promote mesoderm induction (Rosa et al. 1988). However, since these factors do not induce the dorsal mesoderm, they cannot be regarded as “complete” factors. We had also attempted to purify the putative factor from various sources, including calf kidney, mammalian cells, and carp air bladder. During these studies, we found that the conditioned culture medium of K562 cells, a human leukemia cell line, had high mesoderm-inducing activity. A protein isolated from this medium, which retained the mesoderm-inducing activity, was correlated with erythroid differentiation factor (EDF) and contained the inhibin βA subunit, which is also a subunit of activin A (Murata et al. 1988). From these results, it was concluded that the “mesoderm-inducing factor” was in fact activin A (Asashima et al. 1989). After these findings, activin A was identified by many researchers as a mesoderm-inducing factor from several tissues such as XTC-MIF (Smith et al. 1990), mouse myelomonocytic leukemia cells (Albano et al. 1990), mouse macrophage cells (Sokol et al. 1990) and chick embryo (Asashima et al. 1991a; Asashima et al. 1991c). Furthermore, activin A is actually contained in the Xenopus embryo (Asashima et al. 1990, 1991b; Fukui et al. 1993, 1994). From these results, activin A seems to function as a mesoderm-inducing factor in the vertebrate embryo (Asashima 1994; Ariizumi et al. 2001; Okabayashi & Asashima 2003). Our further experiment showed that Xnr5 and Xnr6, which are other members of the TGF-β superfamily and share the same receptors with activin, are also required for mesoderm formation (Takahashi et al. 2000), suggesting that Actvin/nodal signaling plays a crucial role in mesoderm formation in the Xenopus embryo.

Alterations of activin concentration facilitate the induction of various organs and tissues

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

In Xenopus, the ectodermal cell mass in the animal at mid-blastula possesses multipotency, and as such is a convenient material for the so-called “animal cap assay”, which is frequently used for the evaluation of mesoderm-inducing activity by candidate factors. In this assay, a square (measuring 0.6 mm each side) of the animal pole region (animal cap) is dissected and incubated in culture solution (e.g. Steinberg's solution) that contains the candidate factor. This method facilitates the differentiation assay due to the convenience of handling the animal cap. In the absence of any treatment, the animal cap retains its globular shape, and after several days, these cells differentiate into atypical epidermis. If the candidate factor possesses inducing activity, the animal cap is transformed in another cell type. Using this procedure, the molecular properties of activin A have been examined.

Interestingly, we found that activin A could transform animal cap cells into not only mesodermal tissues, but also endodermal and ectodermal tissues other than epidermis (Moriya & Asashima 1992; Kinoshita et al., 1993; Tiedemann et al., 1996; Asashima et al., 1997; Okabayashi & Asashima 2006). This pleiotropic inducing activity is dependent upon the concentration of activin A used (Nakano et al. 1990; Ariizumi et al. 1991; Ariizumi & Asashima 1994; Asashima et al. 1999); summarized in Fig. 1) For example, when animal caps are treated with 0.5–1.0 ng/mL of activin A, they differentiate into ventral mesoderm, which includes mesenchyme and hematopoietic cells. Activin A at 5–10 ng/mL provokes animal caps to induce myotic cell formation, and activin A at 50–100 ng induces notochord formation. Moreover, an activin A dosage greater than 100 ng/mL directs animal caps to become endodermal cells. How do the cells recognize the concentration of activin A? It has been shown that the recognition of activin A concentration is based on the absolute number, not the ratio, of activin receptors occupied by activin A molecules (Dyson & Gurdon 1998).

image

Figure 1. Summary of the in vitro organ induction system using activin A. Animal caps treated with various concentrations of activin A show induction of different types of tissues. By adding retinoic acid (RA), other organs can be induced (orange arrow). Moreover, by changing the conditions of activin A (or RA) treatment, the types of induced organs are increased (green arrow). D&R, Dissociation and reaggregation method; Sd, sandwiching method, in which treated caps are placed between two non-treated caps.

Download figure to PowerPoint

Our experiments have shown that activin A can induce many types of cells, ranging from endoderm to ectoderm. In addition, we could induce other tissue types by treating the animal caps with activin A in combination with other growth factors, given either concurrently or consecutively, or by dissociation of the animal cap. In the following section, we describe in detail the strategies used to induce different organs.

Induction of pronephros by treatment with activin A and retinoic acid

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

As described above, activin A treatment induces animal caps to differentiate into several different types of tissues, although this activity is not universal. To induce the formation of more complex organs and tissues, additional factors are needed. For example, pronephros can be induced by another growth factor, retinoic acid (RA). RA is thought to play a role in anterior–posterior patterning in early embryogenesis. Indeed, treatment of the Xenopus blastula with RA causes posteriorization effects and head formation is severely inhibited. When animal caps were treated simultaneously with 10–4 m RA and 10 ng/mL activin A, pronephric tissues were efficiently induced (Fig. 2A,C; Moriya et al. 1993). This tissue was evaluated by histological sectioning and electron microscopy. Marker gene (Xlim1, Pax2/8, and Wnt4) analysis has shown that in vitro-induced pronephros results in an expression profile that is characteristic of the intact pronephric organ (Chan et al. 2000). Functional acquisition of induced pronephros has been examined in transplantation experiments. When the presumptive kidney region of the tailbud was surgically removed, edema resulted and the tadpole died within 1 week. On the other hand, transplantation of in vitro-induced pronephros allowed the embryo to survive for about 1 month (Chan et al. 1999). In induced pronephros, immunopositivity has been observed for the mouse monoclonal antibodies 3G8 and 4A6, which are specific for the duct and tube, respectively (Osafune et al. 2002). These results suggest that treatment with activin A and RA induces the formation of both the pronephric tube and duct in animal caps. Furthermore, we isolated novel genes involved in kidney development with these induction systems (Uochi et al. 1997; Uochi & Asashima 1998; Sato et al. 2000; Satow et al. 2002, 2004; Kaneko et al. 2003; Kyuno et al. 2003; Li et al. 2005)

image

Figure 2. (A) Strategy for inducing pronephros and pancreas formation with activin A and retinoic acid (RA). Treatment with both activin A and RA induces pronephros (upper column), whereas RA treatment after 4 h incubation with activin A solution causes the animal caps to differentiate into pancreatic tissue (lower column). (B–D) Histological section of animal caps stained with hematoxylin–eosin. (B) Non-treated cap. (C) Animal cap treated simultaneously with activin A (Act) and RA. (D) Animal cap treated with Act and RA in a “time-lag” manner. pro, Pronephros; pan, pancreas; int, intestine.

Download figure to PowerPoint

Time-lag treatment with activin A and RA induces pancreas formation from the animal cap

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

When animal caps were treated with activin A and RA, pancreatic tissues were also induced. However, this combined treatment showed low efficiency in terms of inducing pancreas formation. A “time-lag” method was attempted (Fig. 2A; Moriya et al. 2000a), in which animal caps were treated with 10–4m of RA immediately following treatment with 100 ng/mL activin A. As a results, the rate of pancreas induction was low (about 40%) and the pronephric organ was highly induced (more than 60%; corresponding to pronephros induction). On the other hand, when animal caps were treated with RA 3–5 h after activin A treatment, the rate of pancreas induction was dramatically increased (more than 80%) and pronephros induction was decreased (about 10%). Histological sectioning showed that these tissues had acnus-like structures, and under the electron microscope these explants were found to contain electron-dense granules that are characteristic of exocrine pancreatic tissues (Fig. 2D). Moreover, immunohistochemistry and reverse transcription–polymerase chain reaction (RT–PCR) analyses showed that a pancreas-specific marker was expressed in the animal cap (Moriya et al. 2000b). Using this induction system, we screened for genes involved in pancreas formation (Sogame et al. 2003). Several genes that are specifically expressed in the developing pancreas could be identified.

Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

Using the dissociation–reaggregation method, we could induce cardiovascular organ formation (Fig. 3A). Briefly, 5–10 animal caps were dissociated by incubation in calcium-free medium, treated with more than 100 ng/mL of activin A, and then reaggregated (Fig. 3B). After cultivation for 3 days, cardiomyocyte pulsation was observed in the reaggregate (Fig. 3C). This induced cardiac tissue expressed marker genes, such as Nkx2.5 and CarTI, and was histologically identical to endogenous heart tissue. When the induced heart was transplanted into an embryo from which the cardiac primordia had been surgically removed, the embryo developed into a frog (Ariizumi et al. 2003). Moreover, when the induced heart was transplanted into the abdominal region of a normal embryo, this embryo developed into an adult that had two independent hearts. Interestingly, both hearts were smaller than the heart of a normal frog, although the total volumes appeared to be equal, which suggests that the induced heart shares functions with the endogenous heart. Taken together, these results show that the induced heart functions normally. Utilizing these induced cardiac tissues, many genes involved in heart formation can be isolated (Ito et al. unpubl. data, 2008; M. Yamagishi et al., unpubl. data, 2008).

image

Figure 3. (A) Strategy for inducing cardiac organs by the dissociation–reaggregation method. (B) Reaggregated cap 12 h after reaggregation. (C) One week after treatment, the reaggregate differentiates into a heart structure with periodic pulsation.

Download figure to PowerPoint

Induction of the eye structure from an animal cap

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

In a previous study, it was shown that the anterior neural structure, which includes the forebrain and eye vesicle, can be induced in animal caps by treating with RA and concanavalin A (ConA) (Moriya et al. 1998). In addition, by sandwiching the dorsal lip region and cells of the ventral marginal zone with untreated animal caps, the eye structure could be efficiently induced (Sedohara et al. 2003). Histological examination has shown that although it is not completely identical, the induced eye possesses morphology that is typical of the normal eye, including the lens and retina.

The induced eye contained elongated optic nerves, which connected to the forebrain and recognized an external light stimulus, suggesting that the induced eye has at least partly normal visual function. Nevertheless, it is not easy to induce eye vesicle formation from animal caps treated with activin A. In a recent study, we succeeded in inducing eye vesicle from animal caps treated with activin A using the sandwich method (K. Suzawa et al. unpubl. data, 2008).

Induction of other organs and tissues by activin A

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

Using activin A, we were able to induce the formation of other tissues, such as cartilage, blood cells, vascular cells and blood islands. When animal caps were treated with activin A, precultured for 1 h, and then sandwiched between untreated ectoderm, cartilage-like cells were strongly induced (Fig. 1; Furue et al. 2002; Myoishi et al. 2004). In these explants, the expression of the collagen type 2 and Cart-1 genes, both of which are expressed in chondrocytes, was observed. Furthermore, we could induce erythrocyte or leukocyte with activin and stem cell factor (SCF) or IL-11 (Miyanaga et al. 1998), duct-like structure expressed Flk-1 with VEGF (Yoshida et al. 2005), and blood vessel with angiopoietin (Nagamine et al. 2005). With this system, we isolated novel cardiovascular-related genes (Inui & Asashima 2006).

A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

In a classical transplantation experiment with newt embryos, it has been shown that dorsal lips dissected from the early gastrula induce head structure formation, whereas dorsal lips from the mid-late gastrula induce formation of the trunk-tail structure. This indicates that the organizer region is able to distinguish between the “head organizer” and “trunk-tail organizer”. In molecular analyses conducted in the 1990s, the head organizer was shown to require several factors, such as Cerberus, frzb, and Dickkopf, for head-inducing activity (Bouwmeester et al. 1996; Leyns et al. 1997; Wang et al. 1997; Glinka et al. 1998). Interestingly, head organizer activity could be mimicked by activin in animal cap cells (Ariizumi & Asashima 1995a,b; Ninomiya et al. 1998). When animal caps treated with 100 ng/mL activin A and cultivated for 12 h in saline were sandwiched between two untreated caps, the sandwiched explants differentiated into trunk-tail structures that contained the spinal cord, notochord, and somite. On the other hand, when the incubation time in saline was extended to 24 h, the sandwiched explants differentiated into head structures with eye vesicles and cement glands. This experimental result shows that activin A can play roles in the induction of both the trunk-tail organizer and head organizer.

Induction of tissues from mouse embryonic stem cells

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

Following the establishment of a successful tissue-induction system with amphibian undifferentiated cells, we tested the application of the induction protocols to mammalian stem cells, such as murine embryonic stem (ES) cells. ES cells can be established from the inner cell mass of the mammalian preimplantation embryo. ES cells have generated enormous interest not only for regenerative medicine, but also as a unique model for in vitro differentiation of various organs and tissues. This is due to the capacity of ES cells to self-renew and their pluripotency to differentiate into many different cell types. ES cells have been demonstrated to differentiate into all cell types following transplantation into embryos during early development (Evans & Kaufman 1981; Martin 1981; Doetschman et al. 1985). To date, murine ES cells have been demonstrated to differentiate into various cell lineages in vitro, such as neurons, cardiomyocytes, pancreatic β-cells, and hepatocytes (Loebel et al. 2003). In the remainder of this paper, we describe some examples of how our in vitro tissue-induction methods can be applied to murine ES cells. Our success in this area suggests the existence of differentiation mechanisms that are common to mammalian stem cells and amphibian undifferentiated ectodermal cells.

Induction of cardiomyocytes from murine ES cells

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

Cardiomyocytes are one of the cell types that can spontaneously differentiate from multicellular aggregates of ES cells, which are called embryoid bodies (EB). EB have a similar morphology to early postimplantation embryos, which contain the three primary germ layers. Although the periodic beating characteristic readily distinguishes differentiated cardiomyocytes, efficient inductive methods for cardiomyocytes from ES cells have not yet been established. To date, certain growth factors, chemicals, and feeder cells that promote the differentiation of ES cells into cardiomyocytes have been reported (Johansson & Wiles 1995; Wobus et al. 1997; Sachinidis et al. 2003; Honda et al. 2005). In most of the cases described above, the differentiation of cardiomyocytes has been examined in the presence of serum. For application to regenerative medicine, the establishment of in vitro inductive methods without serum and feeder cells is required, to avoid contamination of the cardiomyocytes from unidentified viruses in the serum or feeder cells. The establishment of induction conditions with defined media and factors would also increase our understanding of the precise regulatory mechanisms of cardiomyocyte differentiation in vitro. However, it is noteworthy that under serum-free conditions, ES cell aggregates do not form the three germ layer-like structure. Therefore, the induction of the mesodermal structure is strongly suppressed (Johansson & Wiles 1995; Watanabe et al. 2005).

To establish induction methods for cardiomyocytes under serum-free conditions, we initially tested activin A as an inducer of this tissue, as we have demonstrated in the frog system. However, regardless of the concentration used, activin A failed to induce cardiomyocytes from murine ES cells. We also tried another TGF-β family member, bone morphogenetic protein 4 (BMP4), for this purpose, since BMP4 has been shown to induce ventral mesoderm formation in Xenopus laevis and from murine ES cells (Dale et al. 1992; Johansson & Wiles 1995; Ng et al. 2005). EB prepared from ES cells were cultured in GMEM that was supplemented with serum replacement (knockout serum replacement, KSR), treated with 1 ng/mL BMP4 for 5 days, and then cultured without BMP4 for the analysis of their fate. After 3 days of culture without BMP4 (on differentiation day 8), spontaneously beating cardiomyocytes appeared in part of the EB, and the numbers of these cells increased progressively over several days of culture. Under these conditions, cTnT-expressing cells were detected uniformly in the EB (Fig. 4A). RT–PCR analyses confirmed the specificity of this cardiomyocyte differentiation, as the transient induction of early mesodermal markers, brachyury and goosecoid, was followed by the expression of lateral plate and paraxial mesodermal markers, such as Flk1, Pdgfrα, and Pdgfrβ. Furthermore, the expression levels of neural markers were all suppressed in the presence of BMP4, which suggests that EB cultured under these conditions specifically differentiate into mesodermal lineage cells.

image

Figure 4. (A–C) Immunofluorescent staining of tissues that differentiate from mouse ES cells. (A) Cardiomyocytes induced in the presence of BMP4 in serum-free medium were stained with anti-cardiac troponin T (cTnT) antibody (green). (B) Pancreatic tissue induced by activin A and RA was stained with anti-insulin C peptide antibody (green), anti-amylase antibody (red), and 4',6-diamidino-2-phenylindole (blue). (C) Ciliated epithelial cells differentiated in serum-free medium were stained with anti-β-tubulin IV antibody (red).

Download figure to PowerPoint

To understand the mechanisms of cardiomyocyte differentiation, specific cell surface markers can be used as powerful tools, since EB and differentiated cells generally comprise heterogeneous cell populations. Flk1, which is also known as vascular endothelial growth factor receptor 2 (VEGFR2), is considered to be a cell surface marker for the progenitors of cardiomyocytes (Motoike et al. 2003; Ema et al. 2006). Purified Flk1-positive cells have previously been demonstrated to differentiate into cardiomyocytes or hematopoietic cells in vitro (Nishikawa et al. 1998; Iida et al. 2005; Yamashita et al. 2005). On the other hand, other researchers have claimed that cells of the cardiomyocyte lineage can be derived from an Flk1-negative subpopulation (Kouskoff et al. 2005). In our search for new cell surface markers for cardiomyocyte progenitors, we have found that N-cadherin is useful for the fractionation of cardiomyocyte progenitors. N-cadherin is continuously expressed in the cardiogenic mesoderm at various stages of mouse development (Radice et al. 1997). Moreover, when the cardiomyocytes differentiated by our protocol were fractionated by fluorescence-activated cell sorting (FACS), the N-cadherin-positive cells showed significantly higher levels of transcripts for the cardiogenic markers Nkx2.5, Tbx5, and Isl1. Moreover, these cells showed a significantly (eightfold) higher ability to differentiate into cardiomyocytes when they were differentiated into stromal cells for 6 days. Therefore, we propose N-cadherin as a novel cell surface marker for the isolation of cadiogenic progenitor cells from in vitro-differentiated cells.

RA has been suggested to regulate cardiomyocyte differentiation. Retinoid X receptor (RXR) α null mice die because of myocardial malformation (Kastner et al. 1994; Gruber et al. 1996), while RA receptor knockout mice show no obvious defects. Although RA signaling is obviously essential during cardiovascular development, it remains unclear how it influences the differentiation of cardiomyocytes. Moreover, RA is spontaneously converted from all-trans RA to 9-cis RA. This chemical instability makes it difficult to evaluate the specific functions of RA in cardiomyocyte differentiation (Pijnappel et al. 1993; Sucov & Evans 1995). Indeed, studies of in vitro differentiation of murine ES cells have demonstrated that both all-trans RA and 9-cis RA can increase the number of cardiomyocytes when EB are cultured in the presence of serum (Wobus et al. 1991; Wobus et al. 1997).

To clarify the influence of RA signaling on cardiomyogenesis, we used RXR-specific agonists and antagonists in the analysis of cardiomyocyte differentiation of murine ES cells under serum-free conditions in vitro (Honda et al. 2005). The number of beating cardiomyocytes that differentiated from murine ES cells increased significantly following treatment with the synthetic RXR agonist PA024. In contrast, when EB were treated with the RXR antagonist PA452, the number of beating foci was decreased in a dose-dependent manner. Our results clarified that RXR signaling regulates in vitro cardiomyocyte differentiation from murine ES cells and probably acts in a similar way during normal development.

Induction of pancreatic tissues from murine ES cells

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

There have been no reports on the induction of whole pancreas or pancreatic tissues from murine ES cells, while several studies have focused on the differentiation of pancreatic β-cells from ES cells (Soria et al. 2000; Lumelsky et al. 2001; Hori et al. 2002; Shiroi et al. 2002; Blyszczuk et al. 2003; Kim et al. 2003; Leon-Quinto et al. 2004). Recently, using an approach similar to that used for the induction of pancreas from the animal caps of Xenopus embryos (Moriya et al. 2000a), we succeeded in differentiating from murine ES cells a pancreas-like tissue, which contained endocrine cells, exocrine cells, and duct-like structures (Nakanishi et al. 2007). With our protocol, EB treated with RA and activin A were differentiated into a complex and functional pancreas that included all the endocrine (α, β, γ, and δ) cells, acinar cells, and pancreatic duct-like structures (Fig. 4B). Moreover, modulation of the concentrations of RA and activin A in this method changed the ratio of exocrine to endocrine cells in the induced tissues.

In the differentiation of pancreatic tissues, RA is required for the induction of pancreatic markers, such as amylase 2, insulin II, glucagon, Pdx-1, and Ppy. In contrast, activin A has been reported to induce endoderm differentiation from both human and murine ES cells (Kubo et al. 2004; D'Amour et al. 2005; Yasunaga et al. 2005). We confirmed this activity of activin A by observing a significant increase in the expression of the early endoderm development marker Sox17 in EB treated with activin A. Interestingly, a low concentration of activin A (10 ng/mL) induced a much higher amylase 2 gene expression level than that found in untreated EB, while at a higher concentration of activin A (25 ng/mL), the expression of insulin was markedly increased in the EB treated with both activin A and RA. These results suggest that activin A plays an additional cooperative role with RA in the differentiation or proliferation of pancreatic endocrine and exocrine cells. The relatively simple system that we have developed for inducing the differentiation of ES cells into pancreas-like tissue in vitro may provide a good model for analyzing the mechanisms of pancreatic development.

Induction of ciliated epithelial cells from ES cells

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

We also established a unique protocol for inducing epithelial cells with cilia. With our method, ES cell aggregates (1000 murine ES cells per aggregate) were prepared by floating culture for 3 days, followed by adhesion culture for 3 weeks on gelatin-coated dishes in Dulbecco's modified Eagle's medium supplemented with 10% KSR. Initially, during our investigation of various induction conditions, we occasionally observed unusually active movement of floating debris derived from some dead cells in the culture medium. This was due to the weak local flow of the medium over the surfaces of differentiated ciliated cells. Observation at high magnification identified the presence of rapidly moving cilia on the surfaces of the epithelial cells (Fig. 4C). To confirm these observations, we captured the ciliary movement using a high-speed camera, and found that the frequency of beating was 17–20 Hz, which is comparable to that of ciliated cells of the normal respiratory tract (approximately 20 Hz). Moreover, electron microscopic observation revealed that the processed microtubules were arranged in the 9 + 2 structure, which is the same specific alignment observed in normal ciliary microtubules. These differentiated cells expressed a ciliary marker protein, β-tubulin IV, as well as hepatocyte nuclear factor-3/forkhead homologue 4 (HFH-4), which is a transcription factor that is essential for ciliogenesis. The differentiated EB that contained these ciliated cells expressed respiratory marker genes, such as those that encode thyroid transcription factor-1 and surfactant protein-C, although there was no expression of Ovgp1, which is a marker gene for the oviduct. Thus, we propose that the ciliated epithelial cells induced in our in vitro study most closely represent those found in respiratory tissues, such as the trachea or lung in vivo, although we could not rule out the existence of ciliated cells from other organs (Nishimura et al. 2006).

With our induction method, the addition of fetal bovine serum to the medium strongly inhibits the induction of ciliated cells from EB. BMP, but not activin A, also inhibit this differentiation. During lung bud extension in early embryonic development, cells that express high levels of BMP4 differentiate into distal cells, whereas cells with low expression of BMP4 differentiate into proximal cells, including ciliated cells and clara cells (Weaver et al. 1999). Therefore, our result is consistent with the model that BMP4 determines the distal–proximal polarity of the respiratory tract. Previous methods for inducing ciliated cells in vitro from embryonic or adult tissues used culturing at an air–liquid interface (Coraux et al. 2005). The air–liquid interface method mimics the condition of the adult trachea, which is different from the conditions under which ciliated cells differentiate during normal development. Our induction system more closely mimics the normal development of ciliated cells, and thus is a useful tool for studying the differentiation mechanism of normal ciliated epithelial cells. In addition, the system could be used to assay the effects of harmful substances on fetal ciliated epithelial cells.

Chromatin-related proteins involved in pluripotent mouse ES cells

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References

In this paper, we have described the various induction effects of activin A and related factors on the differentiation of undifferentiated cells, such as Xenopus animal cap cells and ES cells. The regulatory mechanisms for the multipotency of these cells are quite intriguing. For the maintenance of pluripotency of murine ES cells, some factors, such as leukemia inhibitory factor (LIF), BMP, Oct-3/4 and Nanog, have been described (Pan & Thomson 2007). To identify other proteins that are involved in the regulation of pluripotency, we have performed a differential proteomic analysis of stem cells using protein samples extracted from the pluripotent and differentiated murine ES cells.

Among more than 50 proteins that have been identified as being specifically expressed in the undifferentiated state of stem cells, we have found that chromatin-related proteins are the major proteins that are expressed in the nuclei of pluripotent stem cells. (Kurisaki et al. 2005). Real-time RT–PCR analysis has confirmed that enrichment of some of these proteins in undifferentiated ES cells is also regulated at the transcriptional level. These results suggest that specific chromatin-related proteins are involved in maintaining the unique properties of pluripotent ES cells. Our findings support the idea that chromatin dynamics is important for the regulation of stem cell differentiation and early lineage decisions during embryogenesis (Rasmussen 2003).

In the present paper, we have described how activin A was identified as the mesoderm-inducing factor. We have also introduced examples of in vitro induction of organs from undifferentiated animal cap cells by treatment with various concentrations of activin A. These induction protocols have been further extended to the differentiation of mouse ES cells in vitro. These model systems for the development of various differentiated cell types, tissues, and even whole organs provide excellent tools for the study of developmental processes. Furthermore, they may be useful not only in regenerative medicine for the development of therapeutic options for a variety of diseases, but also for the development of in vitro assay systems for screening various drugs.

References

  1. Top of page
  2. Abstract
  3. Activin as a mesoderm-inducing factor
  4. Alterations of activin concentration facilitate the induction of various organs and tissues
  5. Induction of pronephros by treatment with activin A and retinoic acid
  6. Time-lag treatment with activin A and RA induces pancreas formation from the animal cap
  7. Dissociation and reaggregation of the animal cap by activin A can induce the formation of heart tissue
  8. Induction of the eye structure from an animal cap
  9. Induction of other organs and tissues by activin A
  10. A sandwich explant of an animal cap treated with activin A differentiates into a partial tadpole structure
  11. Induction of tissues from mouse embryonic stem cells
  12. Induction of cardiomyocytes from murine ES cells
  13. Induction of pancreatic tissues from murine ES cells
  14. Induction of ciliated epithelial cells from ES cells
  15. Chromatin-related proteins involved in pluripotent mouse ES cells
  16. Conflict of Interest
  17. References