Stem Cells
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online August 3, 2006
Stem Cells Vol. 24 No. 12 December 2006, pp. 2618 -2626
doi:10.1634/stemcells.2006-0148; www.StemCells.com
© 2006 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2006-0148v1
24/12/2618    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Torihashi, S.
Right arrow Articles by Fujimoto, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Torihashi, S.
Right arrow Articles by Fujimoto, T.

EMBRYONIC STEM CELLS

Gut-Like Structures from Mouse Embryonic Stem Cells as an In Vitro Model for Gut Organogenesis Preserving Developmental Potential After Transplantation

Shigeko Torihashi, Masaki Kuwahara, Takunori Ogaeri, Pu Zhu, Masaaki Kurahashi, Toyoshi Fujimoto

Department of Anatomy and Molecular Cell Biology, Nagoya University Graduate School of Medicine, Nagoya, Japan

Key Words. ESC • Embryoid body • Gastrointestinal tract • Development • Transplantation • Interstitial cells of Cajal

Correspondence: Shigeko Torihashi, Ph.D., 1-1-20 Daikouminami Higashi-ku, Nagoya 461-8673, Japan. Telephone: 81-52-719-1344; Fax: 81-52-719-1344; e-mail: storiha{at}met.nagoya-u.ac.jp

Received on March 15, 2006; accepted for publication on July 28, 2006.

First published online in STEM CELLS EXPRESS  August 3, 2006.

   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 Acknowledgments
 References
 
Recently, we reported the formation of gut-like structures from mouse ESCs in vitro. To determine whether ESCs provide an in vitro model of gastrointestinal (GI) tracts and their organogenesis, we investigated the morphological features, formation process, cellular development, and regional location within the GI tract by immunohistochemistry, electron microscopy, and reverse transcription-polymerase chain reaction. We also examined the developmental potential by transplantation into kidney capsules. The results demonstrated that Id2-expressing epithelium developed first, {alpha}-smooth muscle actin appeared around the periphery, and finally, the gut-like structures were formed into a three-layer organ with well-differentiated epithelium. A connective tissue layer and musculature with interstitial cells of Cajal developed, similar to organogenesis of the embryonic gut. Enteric neurons appeared underdeveloped, and blood vessels were absent. Many structures expressed intestinal markers Cdx2 and 5-hydroxytryptamine but not the stomach marker H+/K+ ATPase. Transplants obtained blood vessels and extrinsic nerve growth from the host to prolong life, and even grafts of premature structures did not form teratoma. In conclusion, gut-like structures were provided with prototypical tissue components of the GI tract and are inherent in the intestine rather than the stomach. The formation process was basically same as in gut organogenesis. They maintain their developmental potential after transplantation. Therefore, gut-like structures provide a unique and useful in vitro system for development and stem cell studies of the GI tract, including transplantation experiments.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 Acknowledgments
 References
 
ESCs are pluripotent and possess the potential to develop into any embryonic or adult cell type [1–3]. Recently, it has been reported that mouse ESCs are able to form gut-like structures composed of cells from the three embryonic germ layers, namely, definitive endoderm, splanchnic mesoderm, and neural crests [4, 5]. The morphological features of gut-like structures reveal that they have characteristic cells located within gastrointestinal (GI) tracts, possessing interstitial cells of Cajal (ICCs), which are known as pacemakers and neuromodulators of GI muscles. Smooth muscles of the gut-like structures display physiological features similar to those found in GI muscles in vivo, including spontaneous rhythmical contractions and intracellular calcium oscillations [6]. The formation process of gut-like structures was analyzed with electron microscopy and semithin sections, and it was demonstrated that the fundamental process of their formation was similar to that which occurred in normal gut organogenesis [7]. In addition, recent analysis has shown that crucial transcription factors in endoderm and embryonic gut development were expressed in gut-like structures and that the expression patterns were similar to those of embryonic guts [8].

Previous data on gut-like structures suggest that they act as excellent in vitro models for the examination of gut organogenesis and provide a useful tool for the study of physiological and developmental aspects of the GI tract. However, little is known about the precise developmental properties of gut-like structures as organs form from ESCs. In the present study, we clarified the in vitro developmental process and unique morphological features of these gut-like structures. Positional specifications of the gut-like structures for regional identification along the anterior-posterior axis of the GI tract were also examined. Finally, transplantation of the gut-like structures to the renal capsule was carried out to ascertain their developmental potential.

The results reveal unique characteristic features of gut-like structures along with those common to mouse embryonic guts. The transplantation provides evidence that the developmental potential of gut-like structures are committed from early stages of development and are stable even if they are transplanted to an in vivo environment. Thus, this new model provides a unique and useful system for basic investigations of GI tracts and also shows that ESCs have a potential as a source for cellular transplantation.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 Acknowledgments
 References
 
ESC Culture
The culture system has been previously described elsewhere [5]. Briefly, murine ESCs (G4-2) (carrying the enhanced green fluorescent protein [EGFP] gene under the control of cytomegalovirus/chicken ß-actin promoter) were expanded in ESC medium (Dulbecco's modified Eagle's medium containing 10% fetal bovine serum [FBS]) with 1,000 U/ml leukemia inhibitory factor (LIF) (Chemicon, Temecula, CA, http://www.chemicon.com). High-quality FBS is very important in the ESC culture. ESCs were then dissociated with 0.25% trypsin and cultured in hanging drops in the absence of LIF. Approximately 500–1,000 ESCs were incubated in a drop with 15 µl of ESC medium without LIF for 6 days (the number of ESCs in a drop depends on their growth rate [5]), and the resulting EBs were plated onto dishes for outgrowth. The stage when EBs were transferred to tissue culture dishes was designated EB0. EBs and gut-like structures from EB2 to EB28 were examined using morphological and molecular techniques.

Immunohistochemistry
Tissues were fixed with Zamboni solution, periodate-lysine-paraformaldehyde fixative, or Bouin's fixative. Samples were collected and embedded in Tissue-Tek OCT compound (Sakura Finetechnical, Tokyo, http://www.sakura-finetek.com), and they were processed for frozen sections (6-µm thickness). For whole mount staining, fixed samples were dyed in tissue culture dishes and were mounted with coverslips directly using PermaFluor Aqueous Mounting Medium (Thermo Electron Corporation, Waltham, MA, http://www.thermo.com). Primary antibodies and their detection systems are listed in supplemental online Data 1. For mouse primary monoclonal antibodies, the M.O.M. immunodetection kit (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com) was used. Mouse embryos (BALB/c) at embryonic days (E) 10.5, 13.5, 15.5, and 17.5 and newborn animals were also sampled. The use and treatment of animals followed the Guide to Animal Use and Care of the Nagoya University Graduate School of Medicine. Confocal images were taken using a laser confocal microscope (Zeiss LEM 5 Pascal; Carl Zeiss, Jena, Germany, http://www.zeiss.com).

Electron Microscopy
Gut-like structures and mouse embryos were fixed with 2.5% glutaraldehyde, 1.25 mM/l CaCl2, and 3% sucrose in a 0.05 M cacodylate buffer (pH 7.4) for 3 hours and postfixed with 1% OsO4. Samples were processed for conventional electron microscopy as described previously [9].

Transplantation Under Kidney Capsule
Gut-like structures from EB6 to EB21 were transplanted under the kidney capsule of 7-week-old female severe combined immunodeficient (SCID) mice (C.B.17/IcrCrj-scid). Three weeks after transplantation, the recipients were sacrificed. The kidneys were fixed with Zamboni solution, and grafted gut-like structures were identified by a florescent dissection microscope (VB-G25 system; Keyence, Osaka, Japan, http://www.keyence.co.jp). They were then embedded in paraffin wax and sectioned for hematoxylin and eosin staining. Some gut-like structures were embedded in OCT compound and processed for immunohistochemistry.

Reverse Transcription-Polymerase Chain Reaction
For gene-specific polymerase chain reaction (PCR), total RNA from gut-like structures was extracted with an RNeasy micro kit (Qiagen, Valencia CA, http://www1.qiagen.com), and total RNA from mouse or mouse embryo was extracted using TRIzol (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). They were treated with RNase-free DNase I (Qiagen) and reverse-transcribed into cDNA. PCR was performed with Blend Taq DNA polymerase and its applied buffer (Toyobo, Tokyo, http://www.toyobo.co.jp). Primers and reaction conditions are listed in supplemental online data 2.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 Acknowledgments
 References
 
Formation Process of Gut-Like Structures
After 6 days of culturing in hanging drops, EBs at EB0 were ovoid and were composed of a small head and a large body. At this stage, some EBs had already developed beating cardiac muscles and attached to the dishes soon after plating. Most EBs began to beat within 2 days after outgrowth (EB2). Development of an EB and the formation of gut-like structures are shown in Figure 1. At EB4 or EB5, cell clusters appeared with prospective epithelial structures in central regions and were considered to be future gut-like structures. At EB10, premature gut-like structures became identifiable by their associated spontaneous weak contractions. They contracted strongly and regularly at EB14 or EB15, and the contractions continued for up to 2 weeks. After this time, period gut-like structures degenerated.


Figure 1
View larger version (69K):
[in this window]
[in a new window]

  		Figure 1. Time-lapse recording of the outgrowth of an EB and development of gut-like structures. At EB2, 2 days after plating onto a culture dish, cardiac muscles usually differentiated and began to beat in the EB (star) (supplemental online Video 1, image Fig. 1 EB2). At EB10, a cell cluster (white arrow) started to contract weakly as a premature gut-like structure. It became larger and showed regular spontaneous contractions (white arrows). After EB22, it began to break down. At EB17, another gut-like structure (black arrow) developed and began to contract. Scale bar = 500 µm.

 
The appearance of an epithelium, smooth muscle cells, and ICCs was demonstrated by immunohistochemistry and compared with gut organogenesis in vivo (Fig. 2). In the embryonic gut epithelium, Id2 was expressed from the period of definitive endoderm before E10.5. The epithelial cells were stratified columnar cells with a narrow lumen and surrounded by Id2-negative mesenchyma (Fig. 2A). At this stage, smooth muscle cells had not yet developed. At E15.5, inner circular smooth muscle layers showed {alpha}-smooth muscle actin immunoreactivity, although the outer longitudinal muscle layer had not yet differentiated. The epithelium shown by Id2 immunoreactivity began to form villi in the small intestine (Fig. 2B). In newborn mice, the epithelium fully differentiated into a single layer of columnar cells and contributed to the long villi and crypts in the small intestine. They were surrounded by inner and outer smooth muscle layers expressing {alpha}-smooth muscle actin immunoreactivity (Fig. 2C).


Figure 2
View larger version (46K):
[in this window]
[in a new window]

  		Figure 2. Comparison of the development between fetus guts and gut-like structures demonstrated by immunohistochemistry. (A–C): Double staining with Id2 (red) and {alpha}-smooth muscle actin (green) on a frozen section of embryonic small intestines at E10.5 (A), E15.5 (B), and newborn (C). At first, Id2 (red) appeared on the stratified columnar cells with a small lumen. At E15.5, smooth muscle expressing {alpha}-smooth muscle actin (green) appeared around the periphery. In newborns, epithelial layer differentiated, and villi and crypts are clearly identified. Scale bar = 100 µm (A) and 20 µm (B, C). (D–F): Whole mount double staining with Id2 (red) and {alpha}-smooth muscle actin (green) in gut-like structures at EB4 (D), EB10 (E), and EB14 (F). At first (EB4), Id2-immunopositive stratified epithelial cells appeared at the center of the cell cluster. Smooth muscle cells with {alpha}-smooth muscle actin immunoreactivity (green) differentiated at EB10. Finally, a single layer of epithelium was completely surrounded by the musculature at EB14, similar to the mouse small intestine at birth. However, the villi and crypts never formed. Some muscle fibers (arrows) did not remain the wall of the gut-like structure and extended to the surrounding tissue attached to the culture dish. Scale bars = 100 µm (D, E) and 20 µm (F). (G–I): Whole mount immunostaining with c-Kit in gut-like structures at EB4 (G), EB10 (H), and EB14 (I). Mesenchymal cell cluster at EB4 shows c-kit immunoreactivity that became heterogeneous at EB10. Finally, at EB14, c-Kit positive cells were recognized to possess multipolar shapes similar to ICCs. This process is similar to differentiation of ICCs at the level of the myenteric plexus in the mouse intestine during embryogenesis. Scale bar = 100 µm (G, H). Abbreviation: E, embryonic day.

 
The formation of the gut-like structures was similar to gut organogenesis. The formation of gut-like structures was examined using whole mount immunohistochemistry and optical slices by confocal microscopy. At EB4, cell aggregations expressing Id2 appeared at the center of undifferentiated cell clusters. They were stratified cells surrounding a narrow lumen (Fig. 2D). At EB10, a smooth muscle layer expressing {alpha}-smooth muscle actin immunoreactivity began to differentiate at the periphery and surrounded Id2-positive epithelium with a distinct lumen (Fig. 2E). Finally, at EB14, the gut-like structures were composed of a single columnar epithelium surrounded by a muscle layer showing {alpha}-smooth muscle actin immunoreactivity (Fig. 2F). The division of inner and outer muscle layers was ambiguous. Sometimes, muscle fibers did not remain in the wall of the gut-like structure and extended widely to the surrounding tissue growing on the culture dish (Fig. 2F). In these cases, we observed contractions of the gut-like structure accompanied by the movement of the surrounding tissue.

Development of ICCs was also similar to that of the mouse embryo in vivo. c-Kit-immunopositive cells aggregated in the mesenchymal cell cluster at EB4 (Fig. 2G). They differentiated into either c-Kit-positive or c-Kit-negative cells at EB10 (Fig. 2H). By whole mount preparation, developed gut-like structures at EB14 were surrounded by c-Kit-positive cells with cellular processes showing multipolar cell shape, as well as ICCs (Fig. 2I). The developmental course was parallel to the differentiation of ICCs in the embryonic gut, as reported previously [9]. Although ICCs were located in the wall of the gut-like structure revealed by whole mount preparation (Fig. 2I), typical two-dimensional distribution pattern shown by ICCs at the level of the myenteric plexus in vivo was not clearly identified.

Characteristic Features of Gut-Like Structures
Gut-like structures showed a large variety of shapes. Many were balloon or dome-like structures with a large cavity, and some were flat with a narrow lumen. Sometimes, two or more domes were fused to form twins. Tubular structures were rarely observed (Fig. 3A–3D). Their sizes also varied, ranging from 200 to 1,500 µm in diameter. Mature gut-like structures contracted spontaneously, and the frequency of their contractions was temperature-dependent, similar to GI smooth muscles.


Figure 3
View larger version (160K):
[in this window]
[in a new window]

  		Figure 3. Appearance, ultrastructure, and expression of mRNA of gut-like structures. (A–D): Dissection micrographs of gut-like structures showing spontaneous contractions. Many had a balloon or dome-like appearance with wide lumens (A). Some had a more flat shape with narrow lumens (B). Fused or twin types were also common (C). Tubular appearance was rare (D). Scale bars = 100 µm. (E): Electron micrograph of the cross section of the wall. Gut-like structures had a single layer of epithelium surrounded by a connective tissue layer and musculature. Scale bar = 10 µm. (F): Fine structure of columnar epithelial cells. They had short microvilli and sometimes electron-dense granules in the cytoplasm. Scale bar = 1 µm. (G): Goblet cells in the epithelium had mucus granules. Scale bar = 5 µm. (H): Entero endocrine cells located in the epithelium had many electron-dense small granules. Scale bar = 5 µm. (I): ICCs were distributed among SM cells. ICCs had electron-dense cytoplasm with many caveolae and mitochondria. Scale bar = 1 µm. (J): Enteric neurons formed a ganglion surrounded by glia. Cell division of the neuron (arrow) occurred in the ganglion. Scale bar = 10 µm. (K): Reverse transcription-polymerase chain reaction analysis shows the expression of smooth muscle and ICC markers. Lanes 1 and 2 are gut-like structures at EB14 and gut from newborn mouse, respectively. The markers for smooth muscles, such as {alpha}-SMA, {gamma}-enteric actin, SM22, and calponin, were expressed in the gut-like structures as well as the gut from newborn animal. The expression of c-kit—a marker of ICCs—was also detected, although it was weaker in the gut-like structures. Abbreviations: {alpha}-SMA; {alpha}-smooth muscle actin; Ep, epithelium; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; ICC, interstitial cells of Cajal; Mu, musculature; SM, smooth muscle.

 
The walls of the gut-like structures were composed of three distinct tissue layers, corresponding to the epithelium, connective tissue layer, and musculature. The epithelium was constituted of columnar epithelial cells with rare microvilli at their luminal surfaces. Goblet cells and endocrine cells were distributed among the columnar epithelial cells (Fig. 3E–3H). ICCs were scattered among smooth muscle cells and confirmed by electron microscopy and immunohistochemistry for c-Kit (Figs. 2I, 3I). Neurons corresponding to the enteric nervous system were fewer than those observed in mouse GI tracts (Fig. 3J). Reverse transcription (RT)-PCR showed the expression of c-kit and {gamma}-enteric actin, indicating the differentiation of ICCs and enteric smooth muscles. SM22, calponin, and {alpha}-actin were also expressed in the gut-like structures (Fig. 3K).

Regional Identification of Gut-Like Structures
The patterning of the gut-like structures according to the anterior-posterior axis in the GI tract was analyzed by immunohistochemistry on serial sections from 20 gut-like structures using region-specific or characteristic markers. The results are summarized in Table 1. H+/K+ ATPase is a marker of the stomach. Gastrin, GATA4, and Sonic hedgehog (Shh) are dominant in the stomach and proximal small intestine. Intestinal-type fatty acid binding protein (I-FABP) and 5-hydroxytryptamine (5-HT) are mainly expressed in the small intestine. Cdx2 is expressed intensely in the colon and to a lesser degree in the small intestine. Most gut-like structures showed Cdx2 immunoreactivity on their epithelium. Some showed the expression of 5-HT, GATA4, and Shh; however, none expressed H+/K+ ATPase, I-FABP, or gastrin immunoreactivity on their epithelium. Examples are shown in Figure 4. A gut-like structure with a narrow lumen expressed Cdx2 and 5-HT immunoreactivities on the epithelium (Fig. 4A, 4B). Another gut-like structure with a wide lumen showed a heterogeneous pattern. Some areas expressed Cdx2 but contiguous neighboring areas did not, whereas Shh and GATA4 were immunopositive on the serial sections (Fig. 4C–4F). The summarized data (Table 1) suggested that the luminal size or shape did not determine the expression pattern of marker proteins.


View this table:
[in this window]
[in a new window]

  		Table 1. Summary of expression pattern of regional markers

 


Figure 4
View larger version (50K):
[in this window]
[in a new window]

  		Figure 4. Regional identification of gut-like structures by immunohistochemistry and reverse transcription-polymerase chain reaction (RT-PCR). The expressions of seven marker proteins were examined by immunohistochemistry on serial sections from 20 gut-like structures, and data are summarized in Table 1. (A): Example of a gut-like structure with a narrow lumen. One section is shown in (B). Scale bar = 20 µm (A, B). (B): Double staining indicates that the gut-like structure shown in (A) expressed Cdx2 (green) and 5-HT (red; indicated by white arrows) in the epithelium. (C): Example of gut-like structure with a large lumen. Its serial sections are shown in (D–F). Scale bar = 20 µm (C–F). (D): Upper right side of epithelium in gut-like structure expressed Cdx2 immunoreactivity, but other epithelial cells were immunonegative. Serial sections in dotted square area are shown (E, F). (E): Cdx2-immunonegative area indicated by dotted line (D) expressed Shh immunoreactivity on serial section. (F): The same area also expressed GATA4 immunoreactivity on the next serial section. (G): RT-PCR analysis of mRNA expression. Lanes 1 and 3 are gut-like structures at EB14. Lanes 2 and 4 are stomach and small intestine of newborn mice, respectively. Gut-like structures expressed neither H+/K+ATPase {alpha} nor H+/K+ATPase ß. Expression of Cdx2 was also confirmed by RT-PCR. Gastrin, however, was weakly expressed in the gut-like structures. Abbreviations: 5-HT, 5-hydroxytryptamine; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

 
Immunohistochemistry did not show H+/K+ ATPase, and this was confirmed by RT-PCR. On the other hand, gastrin-positive cells were never observed by immunohistochemistry, but RT-PCR analysis detected weak expression, suggesting that duodenums providing a small number of gastrin cells were present in the gut-like structures. Gastrin was highly expressed in the mouse stomach (Fig. 4G).

Innervation and Vasculogenesis
Ganglia in the gut-like structures were poorly developed. Figure 3J shows an electron micrograph of a ganglion with rare nerve fibers. Confocal micrographs stained with vascular acetylcholine transporter (VAChT) demonstrated fine varicose fibers and a small number of ganglia composed of a few neurons (Fig. 5A, inset). These observations indicated that the distribution of ganglia was incomplete. Varicose fibers ran in a disorderly fashion throughout the muscle layer and did not form nerve bundles corresponding to an enteric nerve plexus (Fig. 5B). On the contrary, outside the gut-like structures, there were many neurons, either packed or solitary, showing PGP9.5 immunoreactivity (Fig. 5C). These neurons expressed P75, nitric-oxide synthase, and tyrosine hydroxylase immunoreactivity (Fig. 5D–5F). Therefore, neurons developed in EB in vitro, but their migration into the gut-like structures was incomplete.


Figure 5
View larger version (74K):
[in this window]
[in a new window]

  		Figure 5. Undeveloped innervation and default vasculogenesis of the gut-like structures. (A): VAChT immunohistochemistry by confocal microscopy shows at most eight ganglia (arrowheads) in a gut-like structure within a reconstructed three-dimensional image (optical slice, 40 µm thick). Nerve fiber fascicles corresponding to enteric nerve plexus were absent. Scale bar = 50 µm. Higher magnification of a ganglion is shown in inset (upper right). Scale bar = 20 µm (inset). (B): In the musculature, fine varicose fibers expressing VAChT ran irregularly and did not form bundles. Scale bar = 20 µm. (C): Neural marker PGP9.5 shows the development of many neural elements forming ganglia outside gut-like structure (broken line). Therefore, EBs have the potential to develop neurons. Scale bar = 50 µm. (D–F): Immunostaining of neurons around gut-like structures. Expression of p75 (D), NOS (E), and TH (F) immunoreactivity are identified. Scale bar = 20 µm (D–F). (G, H): Flk-1 immunoreactivities indicating vasculogenesis. Embryonic gut at E17.5 (G) showed Flk-1 immunoreactivity, whereas gut-like structures (H) never did. Scale bars = 20 µm (G) and 50 µm (H). Abbreviations: NOS, nitric-oxide synthase; TH, tyrosine hydroxide; VAChT, vascular acetylcholine transporter.

 
Hematopoiesis and vasculogenesis did not occur in the gut-like structures. Although cardiac muscles were identified by their contractions in EBs, neither blood vessels nor blood cells were observed in the structures. Flk-1 immunoreactivity showing vasculogenesis [10] in the embryonic gut in E17.5 was also not detected in the gut-like structures (Fig. 5G, 5H).

Transplantation
To evaluate the developmental potential of the gut-like structures in vivo, they were transplanted under the kidney capsules of SCID mice. Three weeks after transplantation, explants were identified as EGFP-positive grafts under a fluorescent dissection microscope (Fig. 6A–6C). The gut-like structures showed a homogeneous appearance and became 3–5 times as large as their original size. Their growth was restricted to only the capsules where they were inserted. Blood vessels from the host animal invaded the gut-like structures (Fig. 6B, 6D). Nerve fiber fascicles from the host also entered the wall of gut-like structures (Fig. 6D). All transplanted gut-like structures (at EB6, EB15, and EB21; n = 3 each) had lumens but did not develop villi or crypts (Fig. 6E, 6F). Gut-like structures developed smooth muscle expressing smooth muscle actin and ICCs with AIC immunoreactivity [11], respectively, even though premature gut-like structure at EB6 had been transplanted (Fig. 6G, 6H). Cellular components appeared fully differentiated and displayed a mature appearance. They had neither developed smooth muscles or ICCs at EB6 before transplantation as described. Grafts of gut-like structures after completion of their organogenesis at EB15 and EB21 grew and increased their sizes in the kidney capsules. Histological analysis by hematoxylin and eosin staining did not reveal any sign of teratoma. The gut-like structures were maintained for longer periods in the kidney capsule (more than 42 days) without any appearance of degradation (indicated by disruption of the epithelial layer).


Figure 6
View larger version (85K):
[in this window]
[in a new window]

  		Figure 6. Transplantation of gut-like structures into renal capsules of severe combined immunodeficient mice for 3 weeks. (A–C): Fluorescent dissection micrographs showing enhanced green fluorescent protein (EGFP)-positive gut-like structures. Gut-like structure (A) explanted at EB15 showed smooth appearance and remained on surface of host kidney. Gut-like structure explanted at EB21 (B) received blood circulation from the host animal. Invading host blood vessels are clearly observed. Cross-section of the gut-like structure (B) has a lumen (C), and EGFP-expressing transplant never invaded host parenchyma kidney. Scale bars = 1 mm (A–C). (D–F): Hematoxylin eosin staining of the gut-like structures (B, C). Blood vessels (arrows) and a nerve fascicle (arrowhead) (D) supplied by the host entered the gut-like structure. The gut-like structure had a wide lumen (E), as shown in (C), and three distinct tissue layers (F). The epithelial cells differentiated and did not show any sign of the degeneration 3 weeks after the transplant. Scale bars = 20 µm (D–F). (G, H): Immunohistochemistry of the gut-like structure transplanted at EB6. After 3 weeks, the gut-like structure (G) had Id2 (red) immunopositive epithelial cells surrounded by smooth muscle cells expressing {alpha}-smooth muscle actin (green). In the musculature (H), AIC-immunoreactive interstitial cells of Cajal (red) were identified among smooth muscle cells with {alpha}-smooth muscle actin (green). These cell types differentiated after transplantation and had never occurred before explantation. Scale bars = 50 µm (G, H).

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 Acknowledgments
 References
 
The formation process of gut-like structures was investigated previously by electron microscopy and semithin sections stained with toluidine blue [7]. Kuwahara et al. demonstrated the appearance of an epithelium in the center of these cellular clusters, which was followed by differentiation of the surrounding musculature [7]. In the present study, we have confirmed this developmental process using time-lapse microscopy and immunohistochemistry. The cell clusters of these gut-like structures develop an epithelium by Id2, after which they develop an {alpha}-smooth muscle actin layer around the periphery. Transcription factor Id2 is expressed in the intestinal epithelium from the early embryonic stages to differentiate epithelium and suppress tumor formation [12]. After completion of the formation, many columnar epithelial cells and goblet cells and some endocrine cells with 5-HT immunoreactivity have differentiated in the epithelium similar to intestines in vivo. The development of mesodermal cells, such as smooth muscles and ICCs, in the gut-like structures was confirmed by immunohistochemistry and at the molecular level using RT-PCR. Their formation process was compared with that of embryonic gut and was found to be similar, as shown in Figure 2 [9]. The differentiation of smooth muscles and ICCs demonstrated spontaneous contractions, showing physiological properties similar to those observed in vivo [6]. Furthermore, transcription factors required for endoderm formation and gut organogenesis were expressed in gut-like structures [8]. The findings indicate that the gut-like structure is an in vitro physical model of the GI tract. Furthermore, the formation process of gut-like structures is a suitable model to study development of the GI tract in vivo, including basic endoderm-mesoderm interactions.

We recently analyzed the expression pattern of crucial transcription factors in endoderm and embryonic gut development in gut-like structures by in situ hybridization and immunohistochemistry [8]. Sox17, Id2, HNF3ß/Foxa2, and GATA4 were expressed during the formation process of gut-like structures, and the expression pattern was similar to those of mouse embryonic guts. HNF3ß/Foxa2 in the epithelium, a marker of the anterior region of the GI tract, became heterogeneous after EB7 and then disappeared at EB14. We concluded that gut-like structures contain heterogeneous portions of the GI tract. In the present study, the regional identification of gut-like structures, typical of that observed within the GI tract (differentiated characteristically from oral to anal ends), was carried out by immunostaining on the serial sections and RT-PCR analysis. H+/K+ ATPase, an enzyme in the parietal cells and a marker of the stomach [13], was observed neither by immunostaining nor by RT-PCR. A few gut-like structures showed Shh and GATA4 immunoreactivities that are abundant in the stomach and are also observed in the duodenum [14–17]. Most gut-like structures were Cdx2-positive, indicating that they most likely possessed an intestinal or colonic phenotype rather than a stomach phenotype. This is supported by the previous data of HNF3ß/Foxa2 [8]. This in vitro system is therefore liable to form the posterior region of the gut, such as the large intestine, showing Cdx2-expression or 5-HT-positive ileum and jejunum [16, 18–20]. Only a few gut-like structures developed into duodenum expressing GATA4 and Shh tissues. Gastrin, detected in small amounts by RT-PCR but not observed by immunohistochemistry, is also produced in the duodenum in vivo [21]. Heterogeneous expression patterns indicating anterior and posterior regions of the intestine in a single gut-like structure (Fig. 4D–4F) suggest the possible formation from duodenum to colon within one structure, if it were to grow into a longer tube.

Recently, Kim et al. reported that the differentiation of stomach epithelium depended on the inhibition of Wnt signaling by homeobox gene Barx1, and intestinal differentiation represented a default state [22]. This suggests that the gut-like structures in this system show a prototype of the GI tract and that further expression of homeobox genes may play a key role in the differentiation of stomach epithelium.

I-FABP, strongly expressed in the villi of the small intestine in vivo [23, 24], was not detected in the gut-like structures. Furthermore, we did not observe villi within these structures, even in those maintained for a longer periods of time by kidney capsule transplantation. The data suggest that gut-like structures do not correspond to small intestines, if villi formation is an essential criterion for the small intestine. However, crypts were also absent from the gut-like structures expressing Cdx2 strongly and considered to be large intestine precursors. Three-dimensional developments in the epithelium, that is, villi and crypts formation, are not directly controlled by regional specification signals but depend on more local ones, as suggested previously [25–27].

Neurons indicating autonomous or enteric elements were fully differentiated outside the gut-like structures in EBs; however, the nervous system was poorly developed within them, whereas transplanted gut-like structures did not prevent nerve fiber innervation from host tissues. These findings suggest that molecules inducing migration of neuro-precursors into the GI tract, such as glial cell line-derived neurotrophic factor, GFR{alpha}1, ret, endothelin 3, endothelin receptor ß, netrins, and deleted colorectal cancer, were insufficient or completely absent [28–30]. Also, signaling for proliferation or survival of enteric neurons in GI tract, such as Shh and bone morphogenic proteins, could also be disrupted [17, 31, 32]. Thus, this in vitro system of both the gut-like structures and neuro-elements in EBs provided valuable information on the development, differentiation, and migration of the enteric nervous system within the developing gut. Moreover, the in vitro model provides a useful experimental system to investigate GI motility based on smooth muscles and ICCs function and excluding influences from the enteric nervous system.

Although the lamina propria is usually rich in blood vessels, the connective tissue layer in gut-like structures possessed no vascular system consisting of either blood or lymphatic vessels. We have observed neither blood cells nor lymphocytes in the gut-like structures in vitro. This defect in vascularization may be the main reason that the structures degenerated after 1 month in culture. When they were transplanted to renal capsules of SCID mice, they survived longer by nutrition from the host animal through the growth of invading blood vessels. This lack of circulation also suggests an undifferentiated immune system within the clusters.

Previous transplantation reports of mouse embryonic gut to the renal capsule showed that the renal capsule provided a suitable environment for the development of the embryonic gut [33, 34]. When EBs from ESCs were transplanted, they quickly developed various tissues such as teratoma there [34, 35]. Recently, ESCs developing into definitive endoderm with mesodermal cells were also transplanted into the renal capsule [36]. The graft formed a mixture of both endoderm and mesoderm derivatives such as muscles, bones, distal lung epithelia, and gut epithelia. However, they did not develop gut organs composed of three tissue layers. In our experiments, when the gut-like structures after EB6 were implanted, they continued to develop formed mature gut-like structures (Fig. 6G, 6H) that survived for a relatively long time. They neither invaded the kidney parenchyma nor formed multilineage teratomas. The results demonstrated that cell lineages were already decided at EB6, and ESCs sustained their lineages after grafting. Therefore, this system is technologically advanced and suitable to form gut organs in vitro, and it provides new sources for organotypic transplantation experiments.

In conclusion, the gut-like structures are a basic model of gut organogenesis and provide an in vitro assay tool to study the molecular mechanism of cell differentiation and migration.


   DISCLOSURES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
Acknowledgments
 References
 
The authors indicate no potential conflicts of interest.


   ACKNOWLEDGMENTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 Acknowledgments
References
 
We are grateful to Dr. H. Niwa (Riken, Kobe, Japan) for the gift of ESC line G4-2. We also thank to Drs. T. Yamada and M. Takaki (Nara Medical University) for technical advice on cell culture. This work was supported by Research grants-in-aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and by the Hayashi Memorial Foundation for Female Natural Scientists.


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 Acknowledgments
 References
 

  1. O'Shea KS. Embryonic stem cell models of development. Anat Rec 1999;257:32–41.[CrossRef][Medline]

  2. O'Shea KS. Self-renewal vs differentiation of mouse embryonic stem cells. Biol Reprod 2004;71:1755–1765.[Abstract/Free Full Text]

  3. Conley BJ, Young JC, Trounson AO et al. Derivation, propagation and differentiation of human embryonic stem cells. Int J Biochem Cell Biol 2004;36:555–567.[CrossRef][Medline]

  4. Yamada T, Yoshikawa M, Takaki M et al. In vitro functional gut-like organ formation from mouse embryonic stem cells. STEM CELLS 2002;20:41–49.[Abstract/Free Full Text]

  5. Torihashi S. Formation of gut-like structures in vitro from mouse embryonic stem cells. Methods Mol Biol 2006;330:279–285.[Medline]

  6. Ishikawa T, Nakayama S, Nakagawa T et al. Characterization of in vitro gutlike organ formed from mouse embryonic stem cells. Am J Physiol Cell Physiol 2004;286:C1344–C1352.[Abstract/Free Full Text]

  7. Kuwahara M, Ogaeri T, Matsuura R et al. In vitro organogenesis of gut-like structures from mouse embryonic stem cells. Neurogastroenterol Motil 2004;16 (suppl 1):14–18.

  8. Matsuura R, Kogo H, Ogaeri T et al. Crucial transcription factors in endoderm and embryonic gut development are expressed in gut-like structures from mouse ES cells. STEM CELLS 2006;24:624–630.[Abstract/Free Full Text]

  9. Torihashi S, Ward SM, Sanders KM. Development of c-Kit-positive cells and the onset of electrical rhythmicity in murine small intestine. Gastroenterology 1997;112:144–155.[CrossRef][Medline]

  10. Yamashita J, Itoh H, Hirashima M et al. Flk1-positive cells derived from embryonic stem cells serve as vascular progenitors. Nature 2000;408:92–96.[CrossRef][Medline]

  11. Torihashi S, Yokoi K, Nagaya H et al. New monoclonal antibody (AIC) identifies interstitial cells of Cajal in the musculature of the mouse gastrointestinal tract. Auton Neurosci 2004;113:16–23.[CrossRef][Medline]

  12. Russell RG, Lasorella A, Dettin LE et al. Id2 drives differentiation and suppresses tumor formation in the intestinal epithelium. Cancer Res 2004;64:7220–7225.[Abstract/Free Full Text]

  13. Pettitt JM, Toh BH, Callaghan JM et al. Gastric parietal cell development: Expression of the H+/K+ ATPase subunits coincides with the biogenesis of the secretory membranes. Immunol Cell Biol 1993;71:191–200.

  14. Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development 2000;127:2763–2772.[Abstract]

  15. Jacobsen CM, Narita N, Bielinska M et al. Genetic mosaic analysis reveals that GATA-4 is required for proper differentiation of mouse gastric epithelium. Dev Biol 2002;241:34–46.[CrossRef][Medline]

  16. Boudreau F, Rings EH, van Wering HM et al. Hepatocyte nuclear factor-1 alpha, GATA-4, and caudal related homeodomain protein Cdx2 interact functionally to modulate intestinal gene transcription. Implication for the developmental regulation of the sucrase-isomaltase gene. J Biol Chem 2002;277:31909–31917.[Abstract/Free Full Text]

  17. Fu M, Lui VC, Sham MH et al. Sonic hedgehog regulates the proliferation, differentiation, and migration of enteric neural crest cells in gut. J Cell Biol 2004;166:673–684.[Abstract/Free Full Text]

  18. Facer P, Bishop AE, Cole GA et al. Developmental profile of chromogranin, hormonal peptides, and 5-hydroxytryptamine in gastrointestinal endocrine cells. Gastroenterology 1989;97:48–57.[Medline]

  19. Beck F, Erler T, Russell A et al. Expression of Cdx-2 in the mouse embryo and placenta: possible role in patterning of the extra-embryonic membranes. Dev Dyn 1995;204:219–227.[Medline]

  20. Silberg DG, Swain GP, Suh ER et al. Cdx1 and cdx2 expression during intestinal development. Gastroenterology 2000;119:961–971.[CrossRef][Medline]

  21. Kataoka K, Miura J, Takeoka Y et al. Ontogenesis of gastrin cells in the pyloric antrum and duodenum of the mouse. Cell Tissue Res 1985;239:531–535.[Medline]

  22. Kim BM, Buchner G, Miletich I et al. The stomach mesenchymal transcription factor Barx1 specifies gastric epithelial identity through inhibition of transient Wnt signaling. Dev Cell 2005;8:611–622.[CrossRef][Medline]

  23. Cohn SM, Simon TC, Roth KA et al. Use of transgenic mice to map cis-acting elements in the intestinal fatty acid binding protein gene (Fabpi) that control its cell lineage-specific and regional patterns of expression along the duodenal-colonic and crypt-villus axes of the gut epithelium. J Cell Biol 1992;119:27–44.[Abstract/Free Full Text]

  24. Hermiston ML, Wong MH, Gordon JI. Forced expression of E-cadherin in the mouse intestinal epithelium slows cell migration and provides evidence for nonautonomous regulation of cell fate in a self-renewing system. Genes Dev 1996;10:985–996.[Abstract/Free Full Text]

  25. Haramis AP, Begthel H, van den Born M et al. De novo crypt formation and juvenile polyposis on BMP inhibition in mouse intestine. Science 2004;303:1684–1686.[Abstract/Free Full Text]

  26. Abud HE, Watson N, Heath JK. Growth of intestinal epithelium in organ culture is dependent on EGF signalling. Exp Cell Res 2005;303:252–262.[CrossRef][Medline]

  27. Madison BB, Braunstein K, Kuizon E et al. Epithelial hedgehog signals pattern the intestinal crypt-villus axis. Development 2005;132:279–289.[Abstract/Free Full Text]

  28. Young HM, Newgreen D. Enteric neural crest-derived cells: origin, identification, migration, and differentiation. Anat Rec 2001;262:1–15.[CrossRef][Medline]

  29. Kapur RP, Gershon MD, Milla PJ et al. The influence of Hox genes and three intercellular signalling pathways on enteric neuromuscular development. Neurogastroenterol Motil 2004;16 (suppl 1):8–13.

  30. Burns AJ. Migration of neural crest-derived enteric nervous system precursor cells to and within the gastrointestinal tract. Int J Dev Biol 2005;49:143–150.[CrossRef][Medline]

  31. Chalazonitis A, D'Autreaux F, Guha U et al. Bone morphogenetic protein-2 and -4 limit the number of enteric neurons but promote development of a TrkC-expressing neurotrophin-3-dependent subset. J Neurosci 2004;24:4266–4282.[Abstract/Free Full Text]

  32. Goldstein AM, Brewer KC, Doyle AM et al. BMP signaling is necessary for neural crest cell migration and ganglion formation in the enteric nervous system. Mech Dev 2005;122:821–833.[CrossRef][Medline]

  33. Nishijima E, Meijers JH, Tibboel D et al. Formation and malformation of the enteric nervous system in mice: an organ culture study. J Pediatr Surg 1990;25:627–631.[CrossRef][Medline]

  34. Young HM, Ciampoli D, Southwell BR et al. Origin of interstitial cells of Cajal in the mouse intestine. Dev Biol 1996;180:97–107.[CrossRef][Medline]

  35. Chen U, Kosco M. Differentiation of mouse embryonic stem cells in vitro: III. Morphological evaluation of tissues developed after implantation of differentiated mouse embryoid bodies. Dev Dyn 1993;197:217–226.[Medline]

  36. Kubo A, Shinozaki K, Shannon JM et al. Development of definitive endoderm from embryonic stem cells in culture. Development 2004;131:1651–1662.[Abstract/Free Full Text]





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2006-0148v1
24/12/2618    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Torihashi, S.
Right arrow Articles by Fujimoto, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Torihashi, S.
Right arrow Articles by Fujimoto, T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
STEM CELLS THE ONCOLOGIST CME ALPHAMED PRESS JOURNALS