Volume 136, Issue 7, Pages 2074-2091 (June 2009)

21 of 81

ABSTRACT

FULL TEXT

FULL-TEXT PDF (1757 KB)

The Role of the Visceral Mesoderm in the Development of the Gastrointestinal Tract

Received 19 September 2008; accepted 4 March 2009. published online 19 March 2009.

The gastrointestinal (GI) tract forms from the endoderm (which gives rise to the epithelium) and the mesoderm (which develops into the smooth muscle layer, the mesenchyme, and numerous other cell types). Much of what is known of GI development has been learned from studies of the endoderm and its derivatives, because of the importance of epithelial biology in understanding and treating human diseases. Although the necessity of epithelial-mesenchymal cross talk for GI development is uncontested, the role of the mesoderm remains comparatively less well understood. The transformation of the visceral mesoderm during development is remarkable; it differentiates from a very thin layer of cells into a complex tissue comprising smooth muscle cells, myofibroblasts, neurons, immune cells, endothelial cells, lymphatics, and extracellular matrix molecules, all contributing to the form and function of the digestive system. Understanding the molecular processes that govern the development of these cell types and elucidating their respective contribution to GI patterning could offer insight into the mechanisms that regulate cell fate decisions in the intestine, which has the unique property of rapid cell renewal for the maintenance of epithelial integrity. In reviewing evidence from both mammalian and nonmammalian models, we reveal the important role of the visceral mesoderm in the ontogeny of the GI tract.

John P. Lynch and David C. Metz, Section Editors

Abbreviations used in this paper: BMP, bone morphogenetic protein, E, embryonic day, ECM, extracellular matrix, ENS, enteric nervous system, FGF, fibroblast growth factor, GC, glucocorticoid, GI, gastrointestinal, Hh, Hedgehog, IGF, insulin-like growth factor, ISEMF, intestinal subepithelial myofibroblast, PDGF, platelet-derived growth factor, SOCS, suppressor-of-cytokine signaling, VM, visceral mesoderm

Article Outline

• Abstract

• Overview of GI Development

• ISEMFs

• The Development of the GI Smooth Muscle

• Role of the ENS in GI Development

• Development of the Lateral Plate Mesoderm

• Anterior-Posterior Patterning of the Mesoderm and GI Tract

• Mesenchyme-to-Endoderm Signaling in Stomach Development

• Midgut and Hindgut Development

• The Lateral Plate Mesoderm and Mesenchyme in Left-Right Asymmetry of the GI Tract

• The Role of the VM in Radial Patterning, Villus Formation, and the Differentiation of the Epithelial and Mesenchymal Layers

• Transcription Factors That Contribute to Radial Patterning

• Signaling Pathways in Radial Patterning

• Hh: smooth muscle proliferation and positioning of the crypt

• BMPs: smooth muscle development and regulation of the proliferative compartment (crypts)

• Wnts: proliferation versus differentiation

• FGFs

• Glucocorticoid and Thyroid Hormones in GI Development

• Cytokines and Other Signaling Molecules in GI Development

• Conclusions

• Acknowledgment

• References

• Copyright

Inherited or acquired abnormalities of the gastrointestinal (GI) tract cause many chronic diseases in children and adults that often require costly, long-term medical support and organ transplantation. Cell-based therapies are an attractive alternative to solid organ transplantation, but their progress requires a better understanding of the processes that govern cell fate decisions, cell proliferation, and tissue differentiation during GI development. Several aspects of GI development are well conserved from invertebrates to vertebrates, including the role of the visceral mesoderm (VM). Understanding the role of the VM in development may help orient cell-based therapy research.

Overview of GI Development

Development of the vertebrate GI tract, and of the VM in particular, is conserved across species. In very broad terms, it follows this sequence of events: gastrulation, formation of the primitive gut tube from the endoderm, and apposition of the inner leaflet of the lateral plate mesoderm against the endoderm.1 This inner leaflet eventually circles the gut to become the VM (Figure 1). The outer leaflet becomes the somatic musculature.1 During GI morphogenesis, endoderm and mesoderm undergo extensive regionalization, elongation, and coiling. Concurrently, neural crest cells populate the intestine to form the early enteric nervous system (ENS). This period of rapid growth is characterized by regional signals that are largely derived from the mesoderm and pattern the future intestinal domains along 4 axes: anterior-posterior, dorsoventral, left-right, and radial. The mesoderm sends (or receives) instructive or permissive signals to the adjacent endoderm; the instructive signals are sufficient to direct cell fate in the target tissue, whereas permissive signals enable competent tissue to activate a previously “primed” genetic program.1 The most likely scenario in GI patterning is that neither germ layer holds all the information necessary for GI development, but rather that the relative contribution of each tissue varies during ontogeny, with their relative plasticity diminishing as development proceeds.2 Although the mesoderm has an important role in early GI tract morphogenesis, its formation requires signals from the endoderm mediated in part by the Hedgehog (Hh) family of signaling molecules3; this pathway is conserved among species.4, 5, 6, 7, 8

View Large Image
Download to PowerPoint

Figure 1. Early development of the lateral plate mesoderm. Diagram of a cross section through a Xenopus neurula. Dorsal is to the top. The visceral mesoderm (VM, orange) and somatic mesoderm (SM, brown) are 2 leaflets that are in continuity with the somitic (S) mesoderm at the lateral plate. They are situated between the endoderm (yellow) and the ectoderm (e). Endodermal Hh signals produced by the future gut endoderm induce BMP-4 expression in the VM. The ensuing molecular decisions in the VM are shown; BMP-4 activates expression of the transcription factor FoxF1. One of the early actions of FoxF1 is the inhibition of Iroquois (Iro) in the VM, thereby inhibiting the somatic mesoderm program in cells destined to become VM. FoxF1 also induces BMP4 expression. N, notochord; S, somitic mesoderm. Modified and reprinted with permission from El-Hodiri et al.167

Differential regional responses to the Hh signals along the craniocaudal axis lead to the formation of the foregut, midgut, and hindgut domains.3, 5, 9, 10, 11 The foregut gives rise to the esophagus, stomach, duodenum, liver and bile ducts, pancreas, lung, and thyroid. The midgut is the main precursor of the intestine that is distal to the entry of the common bile duct and extends to the proximal transverse colon; this large domain includes the jejunum, ileum, cecum, and ascending colon. Finally, the hindgut evolves into the distal transverse, descending, and rectosigmoid segments of the colon.1, 12 The contribution of mesoderm components to endoderm patterning is well understood for development of the liver and pancreas. The development of these 2 organs, which also depends on instructive and permissive signals from the adjacent mesoderm, is the subject of recent reviews.13, 14, 15

Following morphogenesis, the characteristic GI cell types emerge from the undifferentiated endoderm and the VM differentiates into its specialized components. This phase, which begins around embryonic day (E) 15 in mice and continues postnatally, occurs during metamorphosis in the frog and begins toward the end of the first trimester in humans.16 It coincides with the formation of the GI tract's fifth axis, the villus-crypt axis. This axis is the functional unit of the intestine and is characterized by finger-like evaginations into the lumen, comprising a mesenchymal stalk and an epithelial cover composed of the precursors to the different epithelial lineages. The mesenchyme contains (intestinal) subepithelial myofibroblasts (ISEMFs), both in the stalk and adjacent to the intervillus space, which are important regulators of mesenchymal signaling. In addition, the mesenchyme, also derived from the VM, forms the lamina propria, together with blood vessels, lymphatics, resident immune cells, and the muscularis propria (Figure 2). In the fully mature GI tract, mesodermal derivatives outnumber endodermal derivatives (Figure 2). A brief overview of the development of these different, mesodermally derived components is warranted before examining the molecular network governing VM development.

View Large Image
Download to PowerPoint

Figure 2. Schematic section through the adult GI tract. Each quadrant represents a different segment of the GI tract, from the stomach to the colon; mesenchymal structures are shown in detail. The following structures are conserved along the craniocaudal axis: the lamina propria mesenchyme, the muscularis mucosa, mesenchymal blood vessels, the ISEMFs, the Meissner's plexus, the longitudinal inner and circular outer layers of the muscularis propria, the Auerbach plexus, and the serosa. These structures are represented in all quadrants. The stomach (left upper quadrant) is characterized by an additional oblique inner layer of the muscularis propria. Its epithelium is characterized by deep, bifid glands (g) that fulfill the secretory and endocrine functions of the stomach. In the duodenum (left lower quadrant), the submucosa is recognizable by its abundance of Brunner's glands, which are epithelial in origin. Consistent with the rest of the small intestine, the duodenal epithelium is characterized by the presence of crypts and villi. The other 2 components of the small intestine are the jejunum and ileum (right lower quadrant). The ileum is characterized by the presence of Peyer's patches, which are large lymphoid aggregates in the submucosa. In the jejunum and the colon, the lymphoid aggregates are located in the mucosal tunic rather than in the submucosa. ISEMFs are adjacent to the epithelium lining crypts and villi. Finally, in the colon (right upper quadrant), the epithelium is characterized by straight glands (g) but does not include villi or crypts. The characteristic feature of the colon mesenchyme is the thinner, circular outer layer of the muscularis propria. Modified and reprinted with permission from T. Caceci, DVM (Virginia-Maryland Regional College of Veterinary Medicine).

ISEMFs

As their name suggests, ISEMFs are cells that are located just below the epithelium of both villi and crypts throughout the length of the intestine (Figure 2). ISEMFs are characterized by α–smooth muscle actin expression and are organized in a syncytium between the epithelium and the muscularis mucosa, where they contribute to extracellular matrix (ECM) and basal membrane composition.17 Although it is uncertain whether ISEMFs are derived from the neural crest, mesodermally derived fibroblasts, or smooth muscle cells, they clearly reside within the mesenchymal layer of the intestine at the base of the crypts, next to the muscularis, starting around week 21 of human gestation.18 In mouse embryos, ISEMFs have been described as early as E18.5.19 Similar to epithelial cells, ISEMFs differentiate as they migrate up the crypt to villus axis.20 There is increasing evidence that ISEMFs are major players in GI development.

The Development of the GI Smooth Muscle

The visceral smooth muscle, derived from the VM, is a very large component of the intestine. Its function in the adult intestine is to confer shape and motility to this hollow organ and allow for the advancement of its luminal contents. The muscle layers of the intestine include the circular inner layer and the longitudinal external layer. In the stomach, there is an additional oblique layer (Figure 2). Together, the smooth muscle layers are called the muscularis propria. Mammals have an additional muscularis mucosa that separates the epithelium from the underlying smooth muscle layers and promotes epithelial movement21 (Figure 2). In humans, disorders of enteric myocytes have been described in children with motility syndromes22; a few reports have associated hypoplasia of the muscularis propria with necrotizing enterocolitis or intestinal perforation in infants.23, 24

Little is known about the role of the developing smooth muscle layer during ontogeny. The early muscle progenitors differentiate from the loose mesenchyme that surrounds the primitive gut. The earliest precursor to both types of enteric smooth muscle cells is a common progenitor called the smooth muscle myoblast, which is characterized by the expression of α–smooth muscle actin, as opposed to the more differentiated smooth muscle myocytes, which express α– and γ–smooth muscle actins.25, 26 Formation of the smooth muscle layer progresses in a craniocaudal fashion, starting in the esophagus at E1125 in mice and 50 hours postfertilization in zebrafish.27 This first layer of smooth muscle myoblasts differentiates from undifferentiated mesenchyme to give rise to the circular inner layer. The second, longitudinal, external layer appears in the mesenchyme approximately 24–28 hours later.25, 26 Finally, in mammals, the muscularis mucosa forms subjacent to the epithelium. In each layer, the smooth muscle myoblasts rapidly differentiate into immature smooth muscle myocytes, but their final differentiation into mature myocytes does not occur until after birth.25, 26

Role of the ENS in GI Development

It is beyond the scope of this review to discuss in detail the development of the ENS, which is organized into the submucosal and myenteric plexuses. Each is derived from neural crest cells, which migrate from the neural plate between E9 and E13.28, 29, 30, 31 However, for the purposes of this review, it is relevant to remember that as the gut develops, it is populated by migrating neural crest cells that home to the mesenchymal layers. Genetic manipulation of the developing gut often leads to a Hirschsprung's-like phenotype, suggesting that the integrity of multiple pathways is required for normal innervation and muscle development of the gut.

Development of the Lateral Plate Mesoderm

The molecular network that governs lateral plate mesoderm development depends on signals from the endoderm. Cross talk between mesoderm and endoderm is essential for regional changes during morphogenesis and for cellular differentiation and patterning along the crypt-villus axis.9, 32 This intricate process is the product of tight spatial and temporal control of signaling molecules and transcription factors in both germ layers.

For example, during mouse gastrulation, the early mesoderm is patterned by the adjacent endoderm via secreted signals. Fibroblast growth factor (FGF)-4 is one of the earliest characterized signals33; its role persists after the formation of the gut tube, when it guides the adjacent endoderm along the anterior-posterior axis, inhibiting anterior fates and promoting posterior fates in a concentration-dependent manner.34 The instructive role of FGF-4 is time sensitive, and as development proceeds FGF-4 loses its ability to repress anterior development.34 Shortly after gastrulation, possibly in response to endodermal FGF signals, bone morphogenetic proteins (BMPs) pattern the mesoderm of Xenopus embryos into its different components, including the VM. This example of a dorsoventral morphogen in the mesoderm is illustrated in Figure 3.

View Large Image
Download to PowerPoint

Figure 3. Early specification of mesodermal tissues. Diagram representing the action of BMP-4 as a ventralizing morphogen. In most species, BMP signaling in combination with other pathways patterns the early mesoderm into different cell types.168, 169, 170 In Xenopus, zebrafish, and other species, BMP-4 acts as a mesodermal morphogen that specifies different mesodermal derivatives along the dorsoventral axis: blood, muscle, kidney, and notochord.170, 171, 172, 173, 174 According to its concentration at a given time and place, BMP-4 confers a distinct dorsoventral identity to different parts of the mesoderm.170, 175 In Xenopus and zebrafish, the Vent homeobox-containing genes act downstream of BMP-4 to convert the signaling gradient into distinct cellular responses.176, 177, 178 The essential VM transcription factor FoxF1 is also downstream from BMP-4.36 Although the exact “dose” of BMP-4 that regulates the specification and differentiation of lateral plate mesoderm has not been established, integration of combinatorial regulatory signals is an important paradigm that is repeated throughout development. The Vent pathway of ventral mesoderm specification has been confirmed in zebrafish and amphioxus,170, 179 but mammalian studies are still pending.

Interactions between Shh, BMP-4, and Foxf1 participate in VM morphogenesis and appear to be conserved across species (Figure 1). Studies in chicks have shown that the endoderm Shh signal activates BMP-4 expression in the lateral plate mesoderm5 where Smads (intracellular mediators of BMPs) regulate mesodermal proliferation and later differentiation into smooth muscle.3, 5, 35 In Xenopus embryos, BMP-4 activates the expression of FoxF1, a transcription factor required for lateral plate and VM formation.36 Furthermore, it is sufficient to induce FoxF1 expression in nonmesodermal tissue, and FoxF1 messenger RNA rescues the phenotype of embryos injected with a dominant negative BMP-4 receptor.36 In Drosophila, Dpp (homologous to Smads in vertebrates; see Table 1) signals are both upstream and downstream of biniou, the Foxf1 homologue essential for VM development.37 In mouse, Foxf1 messenger RNA is absent in the foregut structures of Shh−/− embryos.38 Therefore, induction of the VM via Foxf1 depends on endodermal Hh signals and mesodermal BMPs (see Table 1 for species-specific nomenclature).

Table 1.

Gene Nomenclature


Drosophila	Xenopus	Mouse	Chick	Zebrafish	Human
Biniou	FoxF1	Foxf1/Foxf2	Not identified	Not identified	FOXF1
Dpp	Smad	Smad	Smad	Smad	SMAD
Bagpipe	Xbap; zampogna	Nkx 3.2/Bapx1	Bapx1	bapx1	NKX3.2
Tinman	Nkx2.3,2.5	Nkx2.3–2.7	Nkx2.5	nkx 2.3–2.7	NKX 2.3–2.9
vimar	Not identified	Not identified	Not identified	Not identified	Not identified

NOTE. The names of homologous genes in the different species discussed in the text are summarized. The bagpipe family has 2 related but different genes in Xenopus. In vertebrates, tinman has multiple homologues belonging to the Nkx2 cluster. Except for the transforming growth factor β pathway (Dpp/Smad), signaling pathways are not included because molecules referenced in the text have conserved names across species. In the text, when referring to a gene in a general sense, the murine nomenclature is used. Bapx, bagpipe homeobox-containing gene 1 homologue; Fox, forkhead box.

The Drosophila gene biniou and its vertebrate homologues Foxf1 and Foxf2 have emerged as master regulators of VM morphogenesis and differentiation across species.36, 37, 39 In biniou-null Drosophila embryos, VM development arrests before GI morphogenesis is complete37 (Figure 4A and B). Likewise, in Xenopus laevis, FoxF1 knockdown leads to severe defects in gut coiling and elongation, partly because of decreased cell proliferation in the mesoderm36 (Figure 4C and D). In mouse embryos, Foxf1 expression localizes to the interfaces between mesenchyme and epithelium40; knockout studies have shown a role in early lateral plate mesoderm formation and later differentiation. The earliest known function of Foxf1 is in specifying VM (splanchnic) from somatic mesoderm. In Foxf1-null embryos, there is incomplete separation between the visceral and somatic leaflets of the lateral plate; the homeobox gene Iroquois is ectopically expressed in the visceral leaflet, suggesting that Foxf1 specifies VM by inhibiting Iroquois41 (Figure 1). Although Foxf1-null embryos die at midgestation because of defects in the vascular development of the extraembryonic membranes,42 abnormal esophageal development has occurred by this stage.41 Because both alleles of Foxf1 are required for normal VM development, Foxf1+/− embryos also have severe foregut, gallbladder, and lung malformations.38, 43 The closely related gene Foxf2 is differentially expressed during mouse GI development, with expression predominating in the posterior aspect of the developing gut.44, 45 This suggests some functional redundancy between Foxf1 and Foxf2 in the distal mesenchyme, which would explain the predominantly foregut phenotype of the FoxF1+/− mice and distal phenotype of Foxf2−/− mice. Consistent with its expression pattern, Foxf2-null embryos have colonic dilatation and anal atresia, whereas the foregut seems relatively unaffected.46 Although this phenotype is reminiscent of Hirschsprung's disease in humans, no human condition has been associated with absence or mutations of FoxF proteins. In sum, FoxF proteins are essential for early VM development and multiple aspects of intestinal differentiation.

View Large Image
Download to PowerPoint

Figure 4. The role of biniou/FoxF1 is conserved across species. Invertebrate example: Drosophila melanogaster. (A) Expression of VM marker fascilin III (Fas III) in stage 12 Drosophila embryos (arrow). Anterior is to the left. (B) The VM marker Fas III is absent in the biniou mutant, which lacks biniou, the Drosophila orthologue of FoxF1.37 Vertebrate example: effect of FoxF1 loss of function in Xenopus laevis. (C) Ventral view of a normal gut coiling in a 5-day-old Xenopus embryo. (D) Both coiling and elongation are severely impaired when FoxF1 is knocked down using antisense morpholino-oligonucleotides.36 The black cells in C are melanocytes; they are absent in the knockdown embryos. Cranial is to the top.

To further characterize the molecular regulation of early VM development, Jakobsen et al used gene profiling and ChIP-on-Chip (chromatin immunoprecipitation followed by microarray) to identify downstream targets and partners of biniou in Drosophila.39 They showed that although biniou is expressed throughout VM development, its binding to specific enhancer elements is under tight temporal and spatial control. Importantly, they showed that several biniou targets are conserved in mice; the expression of Tcf21, Sall4, and Ptk7 were down-regulated in Foxf1−/− or Foxf2−/− mouse embryos.39 Bagpipe and Mef2 are also important regulators of VM development and differentiation. Each can bind its own set of regulatory elements in the absence of biniou.47 Understanding the differences between biniou-dependent and -independent genes will be an important avenue to explore to advance our understanding of regional and temporal cell fate decisions during VM development. Of course, it will be crucial to confirm these findings in vertebrates.

Unlike some of the other transcription factors known to participate in regional differentiation, biniou/Foxf1 is expressed throughout the length of the developing intestine (Figure 5A and B). This is of interest because there is a wealth of genes known to participate in foregut and hindgut development, whereas there are few midgut candidates. Further studies to address the differential regional regulation of Foxf1 should elucidate whether the molecular network that regulates midgut development is a “default” program that needs to be modified for the development of more recent digestive functions such as the stomach. What factors participate in regionalization of the VM?

View Large Image
Download to PowerPoint

Figure 5. Transcription factors and signaling pathways during anterior-posterior and radial patterning. (A) Early mesoderm formation is under the control of the conserved homeobox genes. Studies in Drosophila have allowed the identification of some early steps in VM formation implicating homeobox genes.180, 181 These include tinman (Nkx2.3-2.9), bagpipe (Nkx3.2), and other genes.181 Tinman is required for mesoderm formation, whereas its downstream target bagpipe promotes VM development.182, 183 Vimar, a downstream factor from bagpipe, further specifies VM development.184 In Xenopus, 2 bagpipe homologues have been identified and their expression has been analyzed, suggesting that these the role of these homeobox genes is conserved in development; Xbap (Nkx3.2) is expressed in the foregut musculature of Xenopus and mouse82, 185 and zampogna in the musculature of the Xenopus midgut, with some overlap in the posterior foregut mesoderm.185, 186 Expression of zampogna in the posterior indicates early anterior-posterior differentiation in Xenopus. Vimar is indicated in green in the figure because it has been identified in Drosophila but not in Xenopus. Foxf1, downstream of Nkx3.2 (bagpipe), is first expressed in the late gastrula and then throughout the VM of the embryo and the adult. Although the exact relationship between these genes and major signaling pathways is incompletely understood, in the early embryo, FGFs, Wnts, and retinoic acid (RA) promote posterior fates while inhibiting anterior fates. Figure of Xenopus embryo at stage 28: © 1994 Pieter D. Nieuwkoop and J. Faber.187 (B) Spatial distribution of selected mesodermal forkhead and homeobox genes involved in vertebrate gastrointestinal differentiation, using mouse nomenclature. Not shown is Nkx2.5, which specifies the pylorus in the caudal segment of the foregut. Anterior is to the left, dorsal to the top. Diagram courtesy of Aaron M. Zorn. l, liver; p, pancreas.

Anterior-Posterior Patterning of the Mesoderm and GI Tract

GI development progresses in a craniocaudal manner, but the mechanisms by which this occurs are not fully understood. The conserved paradigm of GI development, across species, is that Hh ligands are expressed by the endoderm/epithelium and that Hh receptors and targets are expressed by the mesoderm/mesenchyme, which responds differentially according to developmental stage and anterior-posterior position.4, 5, 6, 7, 8, 48 Multiple signaling pathways confer positional cues to the developing intestine (Figure 6). Because of their known role in segmental and anterior-posterior patterning, homeobox gene products are obvious candidates to integrate these positional cues into regional signals in the presumptive intestine (Figure 5A). Detailed expression studies have mapped Hox gene products in the developing gut; the majority are expressed in the mesoderm.49, 50, 51, 52, 53

View Large Image
Download to PowerPoint

Figure 6. Heterotopic recombinations of mesoderm and endoderm. Detailed dissection and recombination experiments in neurula-stage embryos (during regional specification of the primitive gut tube) of Xenopus laevis, the African clawed frog (equivalent to mice with 7–8 somites or a 22-day-old human embryo), have shown that early endodermal regionalization depends on mesodermal signals.188, 189 Although the nature of these signals is incompletely understood, the major signaling pathways appear to be involved. For example, in the chick, the endoderm needs instructive BMP/activin family signals from the lateral plate mesoderm to differentiate64; retinoic acid (RA) in the VM confers a posteriorizing gradient to the developing GI tract.190 In mouse, Xenopus, and zebrafish, mesodermal Wnt signals confer temporally regulated anteroposterior information to the adjacent endoderm.191, 192 Wnt antagonists in the anterior VM are necessary for stomach development (anterior)83 and Wnts probably posteriorize the VM, but this has not been shown. Because this paradigm has been studied in detail in liver and pancreas development, we refer the reader directly to those studies.191, 192, 193 Modified and reprinted with permission from Horb and Slack.189

There are 3 important points about the role of Hox gene products in GI patterning. The first is that their role appears to have been selected for during evolution. Homeodomain-containing genes belonging to the Nkx2 family (Figure 5A) confer positional information.54 Overexpression of vertebrate Nkx2 genes can partially rescue the phenotype of Drosophila embryos null for the tinman gene in the same cascade, indicating a highly conserved function for these genes (tinman is upstream of Nkx2; Figure 5A).55

Another take-home point concerning Hox genes is that they confer regional information that contributes to the formation of boundaries by conveying permissive signals to the mesenchyme.3 In mice carrying a homozygous deletion of the Hoxd3-13 cluster,56 which is normally expressed in the mesoderm, sphincter formation of both the pylorus and anus is absent.56 Furthermore, ectopic expression of Hoxd12 and Hoxd13 causes abnormal or absent ceca.57 Hoxa13−/− mice develop GI stenoses and atresias58; Hoxa13 mutations have been described in humans with limb, GI, and genitourinary abnormalities.59, 60 Misexpression of Hoxd13 in the chick midgut leads to a hindgut phenotype (homeotic transformation). These findings suggest that loss of Hox genes leads to malformation of the intestines of mice, whereas their ectopic expression in chick embryos leads to a homeotic transformation or ectopic expression of region-specific genes. Together with the finding that Hoxa4−/− mice have abnormal colonic musculature61 and Hoxa5−/− mice have abnormal stomach development,62 these findings indicate that Hox gene products probably participate in regionalization in part through the regulation of smooth muscle precursor cell fate and proliferation.

Third, Hox genes likely confer positional information along the mouth-to-anus axis by integrating multiple regulatory inputs. In the early mouse embryo, BMP signaling in the VM is necessary for Hox gene induction; subsequent mesenchymal-to-epithelial signaling controls stomach morphogenesis and lung budding,63 and retinoic acid signaling controls endodermal expression of Hox genes such as Pdx1 (pancreatic and duodenal homeobox 1) or Cdx2 (caudal type homeobox transcription factor 2).64 Cdx genes are regulated by Wnts, and it is hypothesized that the Wnt pathway “co-opted” Hox machinery to execute their anterior-posterior program.65, 66, 67, 68, 69 Further studies are required to determine if this holds true in VM development. Finally, considering that the mesoderm appears to respond differentially to Hh signals along its anterior-posterior axis, Hh might induce mesodermal Hox gene products to induce sphincter formation. This is illustrated by Shh−/− mice that display homeotic transformation phenotype: intestinal transformation of the stomach, annular pancreas, duodenal stenosis, and an imperforate anus. This homeotic transformation of the Shh−/− stomach supports the concept of an Hh-to-Hox pathway, which has now been demonstrated in the foregut and hindgut of chick embryos, mice, and rats.5, 62, 70

Although Shh becomes restricted to the foregut and hindgut after E14.5, Ihh is expressed throughout the length of the intestinal epithelium during development and adulthood.6 These expression patterns are consistent with the phenotypes of knockout animals.6, 7 In contrast to the Shh−/− phenotype, Ihh−/− mice have a phenotype that resembles human Hirschsprung's disease, nearly absent enteric neurons and a thin muscularis propria,71 which is also observed in the Foxf2−/− mice.46 Although an Hh-to-Fox axis has not been explored in detail, the role of Fox genes in this process has been supported by lack of Foxf1 expression in the GI tract of mice with Gli2 and Gli3 mutations.72 Importantly, these phenotypes mimic human syndromes with GI malformations that have been associated with Hh pathway defects, such as Pallister–Hall syndrome73 and the VATER association.74

Mesenchyme-to-Endoderm Signaling in Stomach Development

Stomach development is a unique example of the differential response of the mesenchyme-to-endodermal signals. The stomach of the developing chick has an anterior and a posterior chamber. Recombination experiments of proventriculus (anterior) mesenchyme and gizzard (posterior) epithelium (and vice versa) have shown that the epithelium often adopts the fate of the adjacent mesenchyme.3, 75, 76 However, this is not the case in other parts of the GI tract.77, 78, 79 Signaling by BMPs 2 and 4 participate in this cross talk, although their relative contribution differs according to developmental stage.80, 81 The divergent homeobox-containing gene Bapx1 or Nkx3.2 regulates mesodermal BMP-4 and Wnt5a expression in the chick stomach75 and pylorus morphogenesis in the mouse82 (Figure 6). In mice, the regionally restricted homeodomain-containing protein Barx1 is required for mesenchymal expression of Wnt inhibitors and normal development of the stomach epithelium.83 In chick, mesodermal BMP-4 independently induces expression of Nkx2.5 and Sox 9 in a cell-autonomous manner, leading to the formation of the pyloric sphincter.84, 85 Consistent with the role of BMP in anterior-posterior patterning, constitutive activation of BMP signaling in distal segments of the developing gut mesenchyme leads to ectopic, cell-autonomous expression of the pyloric sphincter marker Nkx2.5.86 In summary, regulation of major growth factors by mesodermal homeobox genes is necessary for epithelial stomach morphogenesis and formation of the pylorus. These observations raise 2 important points. First, regionalization of the developing GI tract is governed in part by circumscribed transcription factors in response to local signals.83 Second, precise regulation of signaling pathways in the stomach suggests that gastric development requires modification of the midgut “default” program. From a developmental perspective, it seems both economical and practical to modify a preexisting program for the development of a new structure with slightly different functional requirements.

Midgut and Hindgut Development

Unlike the stomach, the plasticity of the midgut epithelium, in response to changing mesenchymal signals, appears to be somewhat limited.87, 88 Using heterotopic cross-associations of endoderm and mesenchyme from different segments of the developing rat intestine, Ratineau et al showed that segments of intestinal mesenchyme from rat fetuses did not confer equal regional information to adjacent endoderm. In the 14-day-old fetus, midgut endoderm did not change its enzyme expression when cocultured with proximal or distal intestinal mesenchyme.89 Conversely, colon endoderm did respond to the association with jejunal or ileal mesenchyme by developing into small intestinal epithelium. Therefore, the hindgut endoderm, like its anterior counterpart, shows relative plasticity late in development in response to instructive signals from the VM. ISEMFs appear to have a role in this process; ISEMFs from anteroposterior segments of the intestine secrete varying amounts of growth factors, thereby conferring dose-dependent regional characteristics to the adjacent endoderm.19 Similar experiments in chicks showed that recombination of proventricular mesenchyme with midgut endoderm did not alter epithelial gene expression in the midgut but did in the foregut.90 Taken together, these findings suggest that the midgut endoderm program might be specified earlier than the cranial- and caudal-most segments of the gut.

The Lateral Plate Mesoderm and Mesenchyme in Left-Right Asymmetry of the GI Tract

In all vertebrates, shortly after gastrulation, the lateral plate mesoderm participates in left-right cell fate decisions. The transforming growth factor β family member Growth and differentiation factor-1 (Gdf-1)91 promotes the expression of genes that regulate lateral development, such as nodal, Lefty, and Pitx2 in the left lateral plate mesoderm.92, 93, 94, 95, 96 Recently, events downstream of Pitx2 (paired-like homeodomain transcription factor 2) and its partner LIM-homeodomain containing transcription factor Isl1 (islet-1) were identified in the dorsal mesentery. The dorsal mesentery, which is continuous with the visceral mesenchyme, derives from the lateral plate mesoderm and anchors the gut to the dorsal wall of the vertebrate embryo. Pitx2 and Isl1 control expression of glycosaminoglycans and the ECM protein N-cadherin; expression is asymmetrically distributed in the dorsal mesentery of chick and mouse, leading to changes in cell shape and intercellular connections that induce the first gut tilt to the left.97 This study has shown a role for the mesoderm-derived mesentery in gut morphogenesis. Because Shh−/− and Ihh−/− mice display malrotation, it is probable that endodermal Hh signals are upstream of this pathway.71 In Xenopus, retinoic acid–metabolizing enzymes are expressed in the VM of the gut during the early stages of gut coiling.98 Addition of retinoic acid or retinoic acid inhibitors to embryonic culture medium before the onset of gut coiling leads to abnormal chirality and organ heterotaxy.98 Thus, the Gdf-1/Pitx2/ECM pathway likely results from the integration of multiple signaling gradients in the mesenchyme.

The Role of the VM in Radial Patterning, Villus Formation, and the Differentiation of the Epithelial and Mesenchymal Layers

In addition to its role in anterior-posterior and left-right patterning, the mesoderm is required for radial patterning and villus formation. Radial patterning is the organization of the GI tract into concentric layers.80 Differentiation of the crypt-villus axis happens at different times in development in different species and progresses in a craniocaudal fashion. In mice, villus formation begins around E15, but crypt formation does not occur until the postnatal period. In Xenopus, the trough/intestinal fold axis is the product of metamorphosis, and in humans, the newborn infant has a complete crypt-villus axis. Crypt-villus axis formation relies largely on mesenchymal-epithelial interactions that determine villus length, crypt depth, and spacing between villi.77, 99, 100, 101 The inaugural event in villus formation is mesenchymal condensation, which depends on ubiquitous signaling pathways that vary according to the developmental time period. Certain transcription factors that participate in early morphogenetic processes also contribute to radial patterning and villus formation.

Transcription Factors That Contribute to Radial Patterning

The transcription factors Foxf1/f2 are expressed throughout development; apart from their roles in early patterning of the lateral plate, they also have a role in later differentiation.36, 46 Studies of mouse knockout models have shown that Foxf proteins are required for multiple aspects of VM development, including radial patterning. Foxf1/2 expression in the early VM is ubiquitous and Foxf proteins are required for VM development in multiple species. However, because of the pleomorphic phenotypes of knockout animals, it is difficult to determine the precise role of Foxf genes. Several other transcription factors participate in radial patterning, with their respective roles still incompletely understood. The difficulty in understanding the precise role of a given transcription factor stems from the developmental interdependence of mesenchymal components among themselves and with the adjacent epithelium. Table 2 summarizes the phenotypes associated with loss of function of forkhead or homeobox-containing transcription factors in the VM. Smooth muscle, enteric neurons, and ECM appear to be the most vulnerable components of the developing VM in response to genetic manipulation. In addition to changes in the VM, genetic manipulation of the VM usually leads to aberrant epithelial development and proliferation.

Table 2.

Transcription Factors in Radial Patterning of the Gut


Transcription factor	Model	Expression	Phenotype	References
FoxF1	Xenopus	VM	Absent lumen, absent smooth muscle actin expression	36
Foxf1+/−	Mouse	VM, foregut	Esophageal atresia, tracheo-esophageal fistula	38, 41
Foxf2−/−	Mouse	VM	Abnormal muscularis propria, ECM, absent ENS, dilated colon, increased epithelial proliferation	38, 41
Hlx−/−	Mouse	VM	Epithelial proliferation, ENS abnormalities, short gut	195, 196, 197
Nkx2.3−/−	Mouse, chick	VM midgut and hindgut	Thin mesenchyme, abnormal villus formation	198, 199
Hoxa5−/−	Mouse	VM midgut	ENS, smooth muscle, ECM	62

Signaling Pathways in Radial Patterning

Multiple cell fate and positional decisions contribute to the formation of the organized, tubular gut. To this end, transcription factors regulate and integrate numerous secreted signals, leading to the differentiation of each mesodermal component and its orientation along the crypt-villus or epithelium-serosa axes. Although these different signaling pathways overlap and their relative contribution is still incompletely understood, we will now discuss their individual roles as they are characterized to date.

Hh: smooth muscle proliferation and positioning of the crypt

Loss-of-function studies of different components of the Hh pathway have led to the comparatively well-understood role of this pathway in the radial patterning of the GI tract. In very broad terms, its functions are 4-fold, keeping in mind, however, that these vary according to developmental window. First, Hh ligands act as radial morphogens for the development of the smooth muscle layer; in other words, too much or too little ligand inhibits smooth muscle development, while the right concentration, regulated by the distance from the endodermal source, induces smooth muscle formation in both chick and mouse.35, 71 Second, depending on developmental window, endodermal Hh signals are important for induction of reciprocal signaling from the mesenchyme back to the endoderm, because loss of function leads to epithelial proliferation.102 Third, epithelial Hh ligands influence ISEMF localization along the crypt-villus axis, probably because these cells serve to integrate the epithelial signal, thereby regulating the proliferative effect of the Hh signal.103 Fourth, Shh and Ihh have complementary roles in ENS development; Shh inhibits ENS development, while Ihh is essential.35, 71 In summary, the Hh pathway serves as a radial morphogen affecting both cell fate and proliferative decisions in the layers of the mesenchyme and epithelium. In the future, targeted, tissue-specific modification of Hh signaling to the mesenchyme should reveal the relative contributions of Ihh or Shh to the development of the different concentric layers. Further, it may help elucidate the concentration and temporal needs of each cell type, thereby facilitating the direction of cell fate in vitro.

BMPs: smooth muscle development and regulation of the proliferative compartment (crypts)

BMPs are members of the transforming growth factor β superfamily; BMPs 2 and 4 have important roles during GI development: signaling via the BMP receptor (BMPR)-1a.80, 104 BMP transcription in the VM is initiated in response to endodermal Hh signals,5 so many of the BMP-related phenotypes are reminiscent of Hh loss-of-function phenotypes discussed in the previous section. Although the role of BMPs in GI development also varies according to developmental stage and tissue layer, studies in the early developing gut of avian embryos suggest that BMP signals are required by all gut layers that are distal to the stomach105 and contribute to the morphogenesis and differentiation of all components of the GI tract.106 BMPs, previously discussed for their role in anterior-posterior patterning, are also necessary for radial patterning and villus formation. In simple terms, BMP-4 in the VM controls smooth muscle proliferation and differentiation throughout most of the developing vertebrate GI tract107; this is probably conserved across species. Whereas the small intestine needs sufficient expression of BMP-4 in the mesenchyme for the development of an adequate muscularis propria, overexpression of BMP-4 in the mesenchyme of the embryonic chick gut delays differentiation into smooth muscle106 and inhibition of BMP-4 in the stomach is required for the highly developed musculature to meet its functional needs.106 Absence of BMP-4 from the presumptive hindgut mesoderm promotes aberrant endoderm proliferation and ENS development, indicating the importance of BMP-4 not only in smooth muscle development but also in the regulation of epithelial proliferation.81

BMP signaling is important for villus formation and endodermal/epithelial patterning (Figure 7A and B). Its role in epithelial morphogenesis and differentiation is complex and involves cross talk between epithelium and mesenchyme; several studies have attempted to address this issue.106, 107, 108, 109 When the secreted BMP inhibitor noggin is constitutively expressed in the developing mouse epithelium, an excess number of crypts develop.107, 110 Depending on the model and the developmental stage analyzed, epithelial and mesenchymal proliferation and differentiation are affected, leading to aberrant crypt and villus formation, similar to that associated with the pathogenesis of human juvenile polyposis.110 However, when BMP signals to only the epithelium are specifically blocked using a villin-dnBMPR1a (dominant negative BMP receptor 1a) transgene, only epithelial proliferation and secretory lineage differentiation are affected; villus or crypt number are not.109 Taken together, these studies support a model in which BMP signaling in the mesenchyme sends antiproliferative signals to the crypts and epithelial stem cells, probably through a signaling relay in the pericrypt mesenchyme (possibly ISEMFs), that regulates the size and location of the proliferative compartment.108 It is unclear whether the mesodermal BMP brake is conserved across species and whether it is in effect at all developmental time points. In contrast to the evidence presented from mice, during gut remodeling associated with Xenopus metamorphosis, mesenchymal BMPs promote rather than inhibit epithelial proliferation.111, 112 The obvious next step in elucidating the relative roles of BMP signaling in the epithelium and mesenchyme is the conditional modification of BMP signaling in the mesodermal layer, something impeded by lack of appropriate tools.

View Large Image
Download to PowerPoint

Figure 7. Molecular cross talk between the VM and the endoderm during crypt-villus formation and in the adult. (A) Signaling pathways during crypt-villus formation. Hh ligands signal to the mesenchyme. In response, probably via ISEMFs, mesodermal Wnt ligands promote proliferation in the intervillus space and in the villus (starting at E16.25) and BMP signals regulate proliferation of the epithelium. Arrows indicate direction of signal. (B) In the adult with a fully developed crypt-villus axis. On the left, pathways expressed in the mesenchyme; on the right, pathways expressed in the epithelium. The signals in the adult likely occur simultaneously. Modified and reprinted with permission from Crosnier et al.194

Wnts: proliferation versus differentiation

Wnts are accepted as the master regulators of epithelial proliferation and differentiation in the developing and adult GI tract.113, 114, 115 Most studies have focused on the role of Wnts in endoderm specification and differentiation, because regulation of nuclear β-catenin accumulation is required for epithelial homeostasis and carcinogenesis.83, 116, 117 The origin of these Wnt signals is mostly mesodermal118, 119 (Figure 7A and B). The exact cellular origin of the Wnt signal in vivo is still unclear, but ISEMFs are likely to contribute because they express Wnt ligands and because they are adjacent to the proliferative compartment of the epithelium, the intestinal crypts.118

In situ hybridization analyses of mouse and chick embryos have shown that the mesenchyme of the developing and adult intestine contains high levels of Wnt pathway components, both canonical and noncanonical.119, 120 In the adult, Wnt signals are necessary for maintenance of the proliferative compartment of the epithelium.121 During development, it has been accepted that the intervillus mesenchyme expresses Wnt components because signs of Wnt activity were found in the epithelium.122, 123 However, novel findings suggest the adult pattern of crypt-predominant Wnt activity only begins around postnatal day 3 in mice; before this time point, canonical Wnt activity predominates in the nascent villi coincident with the mesenchymal expression of several Wnt ligands, including Wnt5a119, 124 (Figure 7A). Importantly, in this study, β-catenin activity does not correlate with epithelial proliferation, suggesting that Wnt signaling may be more than just an on/off switch for proliferation versus differentiation.124

Recent studies have highlighted multiple roles for Wnt5a in the mesenchyme.119, 125 Wnt5a is expressed both in the embryonic and adult mesenchyme, suggesting that it is an important signaling molecule in intestinal development and homeostasis, probably with different spatial and temporal functions, acting through both the canonical and noncanonical pathways.119 Wnt5a−/− mice have multiple defects: thinner muscularis propria, improper midgut closure, and a dramatically shortened midgut. This latter effect is in part mediated by defective post–mitotic cell intercalation in the epithelium in a non—β-catenin–dependent manner.125 This finding joins that of others in highlighting a role for the mesenchyme in gut elongation and morphogenesis125, 126 and highlights the importance of Wnt signaling in more than epithelial proliferation and differentiation.

Together with other Wnt ligands, Wnt5a is also essential during radial patterning when it cooperates with Hh, BMPs, and Forkhead genes. As outlined previously, Foxf mutants have altered collagen production and decreased mesenchymal BMP4 signaling, which leads to Wnt5a overexpression and increased epithelial proliferation.46 Similarly, Foxl1 participates in the regulation of Wnt signals to the adjacent epithelium. Although Foxl1 positively regulates BMP-4 and BMP-2 expression, its main mechanism of action involves proteoglycan synthesis in the mesenchyme.127, 128 Ectopic syndecan-1 synthesis in the villus mesenchyme of Foxl1-null mice is associated with aberrant epithelial expression of nuclear β-catenin and increased epithelial proliferation. Histologically, Foxl1-null mice have delayed formation of intestinal villi compared with wild-type mice, but this is not observed in Foxf mutant mice, in which the mesodermal structures appear to be normal.127, 129 Adult Foxf mutant mice have multiple epithelial abnormalities, including increased cell proliferation in the villi, cystic inclusions, and abnormal stomach histology.127, 129 Loss of Foxl1 augments the epithelial effect of adenomatous polyposis coli (APC) germline mutations by increasing the multiplicity of intestinal adenomas.128 The developing intestine expresses Wnt ligands, receptors, and inhibitors, and their roles vary during development. The control of mesodermal Wnt signals is of particular importance because of the contribution of these factors to epithelial tumorigenesis. Tissue-specific modulation of Wnts in the mesenchyme will improve our understanding of the role of Wnts both in the mesenchyme proper and in the adjacent epithelium.

FGFs

The role of the FGF pathway in radial differentiation is less well understood that that of the Hh, BMP, or WNT signaling pathways. In Xenopus, FGF signaling is necessary for expression of the smooth muscle cell marker XSM22α130 and smooth muscle actin and patterning of the early intestine.131 In the E18.5 mouse embryo, FGF13 is highly enriched (47.6-fold) in the mesenchyme compared with epithelium132 and likely mediates autocrine actions via FGFR1 and R2, as well as having a role in crypt morphogenesis via FGFR3133 (Figure 7B). Normal cecum development is characterized by an outward growth of both mesenchymal and epithelial proliferation134 and requires mesenchymal FGF-10 and BMP-4 signaling in response to epithelial FGF-9 signals.134 Mouse studies indicate that epithelial FGF-9 signaling controls gut length by regulating proliferation and differentiation of subepithelial fibroblasts, possibly through putative mesenchymal stem cells.126 FGFs are required for normal development of the gut because they regulate mesenchymal signaling, but we need to learn more about the time periods and regions in which these potent signaling molecules are active or inactive. Like other pathways, the understanding of their relative contribution to mesodermal and epithelial patterning will be vastly aided by the development of mesoderm-specific tools.

Glucocorticoid and Thyroid Hormones in GI Development

Glucocorticoids (GCs) promote maturation and differentiation of the human and rodent intestinal epithelium.87, 135 However, the mesenchyme appears to be essential for this effect. In vitro studies have shown that fetal rat intestinal endoderm cells do not respond to GCs, but when they are cultured with mesenchymal cells and GCs, α-glucosidase expression is induced.136, 137 Furthermore, although explants of intact perinatal colon do not respond to GCs, sucrase expression is induced when explants derived from colonic endoderm are associated with small intestinal mesenchyme and exposed to GCs.138 These findings indicate that small intestinal mesenchyme, but not colonic mesenchyme, enables GC induction of maturation in an associated epithelium. Mesenchymal cells alone respond to GCs; these cells express the GC receptor137 and show increased expression of collagen type IV messenger RNA139 following in vivo administration of GCs.139, 140, 141 In addition, in vitro studies have shown that in response to GCs, mesenchymal cells deposit laminin at the mesenchymal-epithelial interface, which appears to be essential for the effects of GCs on the epithelium because the effects can be blocked in vitro by the addition of anti-laminin antibodies.142 These experiments suggest that the mesenchyme acts as either a direct GC target that signals to the epithelium or that the mesenchyme induces the adjacent epithelium to respond to the hormone.142 Regardless of the specific mechanism, questions of clinical significance are whether the mesenchyme is responsible for the relatively narrow developmental stage during which the intestinal epithelium can respond to GCs143, 144 and whether these mesenchymal effects are permissive or instructive.

Thyroid hormone also has a role in the development of the vertebrate GI mesenchyme. Thyroid-responsive genes are expressed in the mesenchyme of E18.5 mouse embryos,132 and absence of the thyroid hormone receptor T3Rα leads to hypoplastic smooth muscle layers in the developing intestine.145 This finding correlates with increased expression of T3Rα in the intestinal muscle layers of wild-type animals,145, 146 which might suggest that smooth muscle cells require a thyroid signal for development. The role for thyroid hormone in mammalian epithelial maturation has been difficult to assess because changes in thyroid status cause concomitant changes in circulating GCs.87 However, thyroid hormone can synergize with GCs in eliciting developmental changes in the epithelium,147, 148 but it is not known whether the mesenchyme has a part in this synergy. Xenopus species are characterized by a thyroid hormone–driven metamorphosis during which the GI tract undergoes remodeling characterized by extensive changes of the mesenchyme and development of the muscular layers. Although this developmental step is unique to anurans, it is an attractive model to examine intestinal mesoderm/mesenchyme development; in 7 days, the 2-cell-thick VM proliferates into longitudinal and circular muscle layers as well as numerous other cell types.146 Cultures of Xenopus intestinal epithelial cells are insensitive to thyroid hormone–induced apoptosis when they are cocultured with ECM molecules,149, 150 suggesting that the ECM confers survival signals to the epithelium. Therefore, the study of GI remodeling during Xenopus metamorphosis should offer important insight into the development of the VM and its components.151

Cytokines and Other Signaling Molecules in GI Development

In addition to modulators of the major signaling pathways and hormones, the VM of E18.5 mouse embryos is enriched in multiple peptide growth factors and cytokines.132 These include endothelins, insulin-like growth factor (IGF), and platelet-derived growth factor (PDGF). For example, PDGF is believed to regulate maturation of smooth muscle precursors via the Wnt pathway.152 PDGF-A and its receptor PDGFRα contribute to formation of the crypt-villus axis.153 They are both expressed in the developing mouse small intestine starting around E15.5; PDGF-A is expressed in the epithelium, and PDGFR is expressed in the mesenchyme. PDGF-A−/− and PDGFRα−/− mice have similar phenotypes; compared with wild-type mice, they have fewer and broader villi, a thinner mesenchyme, and premature expression of smooth muscle actin in the villus mesenchyme.153 Therefore, PDGF-A must promote cell proliferation and prevent premature differentiation of the mesenchyme. Again, the role of this molecule depends on the developmental stage, because mutations in the PDGFRα are associated with GI stromal tumors in humans.154, 155, 156, 157, 158

IGF-1 is expressed in the mesenchyme of the E18.5 intestines of mice,132 and IGF-1 receptors are expressed in the submucosal region of the neonatal small intestine; expression is down-regulated after birth.132, 159 IGF-1 and IGF-2 binding to the IGF receptor promotes epithelial differentiation.159 Suppressor-of-cytokine signaling (SOCS) 2 is normally required for regulating the effect of IGFs. Consistent with this role, SOCS2-null mice have increased size and weight, intestine length, and thickness of the lamina propria.160 Members of the SOCS family are highly expressed in the E18.5 mesenchyme,132 suggesting that, like other growth factors, IGF and cytokine signaling in the mesenchyme is tightly regulated.

The peptide endothelin-1 and its receptors are also expressed in the E18.5 mesenchyme of mice.132 The colonic mesenchyme of rat embryos expresses high levels of endothelin 3 at E16.5, which promotes differentiation of adjacent epithelia in a dose- and region-dependent manner.161

Epimorphin (encoded by Stx2) is also up-regulated in the mesenchyme,132, 162, 163 where it is secreted by ISEMFs to mediate epithelial morphogenesis and control intestinal length.162, 163 Stx2-null adult mice have longer intestines and colons than their wild-type counterparts.162 Three-week-old Stx2-null pups have increased crypt cell number and increased villus length, suggesting that epimorphin controls cell proliferation during the final stages of crypt-villus axis formation. This effect has been proposed to result from decreased BMP-4 signaling and increased expression of β-catenin target genes.162 It is unclear why loss of the protein confers a regenerative advantage, but therapeutic manipulation of this pathway may be attractive to pursue.

Smooth muscle genes are up-regulated in the mesenchymal compartment,132 and studies in zebrafish have shown that visceral smooth muscle cells are involved in GI ontogeny.27, 164 Zebrafish expressing a mutant smooth muscle myosin heavy chain have abnormal smooth muscle development, defects in formation of the posterior-most segment of the intestine, uncontrolled epithelial proliferation, and aberrant epithelial expression of ECM proteins.164 This phenotype is similar to that of Ihh−/− and Foxf2−/− mice. It is possible that smooth muscle myosin heavy chain is a downstream target of a putative Ihh-Foxf pathway that controls GI radial differentiation.

Matrix metalloproteinase molecules are expressed in the mesenchyme of E18.5 mouse embryos,132 and high levels of matrix metalloproteinases are transcribed in the mesenchyme during intestinal remodeling in Xenopus embryos (during metamorphosis). Matrix metalloproteinases cleave various ECM components and are target genes for thyroid hormone; therefore, they are likely to regulate ECM composition, basement membrane fenestrations, and subsequent signaling to the adjacent epithelium during remodeling.165, 166 Although the role of matrix metalloproteinase in mammalian GI development is poorly understood, amphibian models offer insight into potential mechanisms.

Conclusions

The VM is a complex tissue that gives rise to the smooth muscle, mesenchyme, and other cells that compose the outer layers of the mature intestine and has multiple essential functions during the development of the GI tract. The development of each VM cell type is intricately linked to that of its neighbors, and although many of the molecular interactions are incompletely characterized, abnormal mesenchymal signaling almost invariably leads to aberrant epithelial proliferation. The VM and its components confer instructive and permissive signals to the adjacent developing epithelium; this instructional dominance appears to be mediated by soluble factors and signaling pathways that regulate development of the mesoderm and endoderm.

Compared with the foregut and hindgut, the development of the midgut is less well understood; it responds to different mesodermal signals. The developmental plasticity of different GI segments is controversial, which could be the result of the different models and conditions used in experiments. Alterations in mesodermal gene expression patterns lead to abnormal mesoderm differentiation but also to epithelial proliferation, so one of the major roles of the VM appears to be control of epithelial proliferation. Studying the molecular basis of GI developmental plasticity and the control of epithelial proliferation by the mesenchyme could identify new therapeutic targets for cancer and other diseases of the GI tract. Improving our knowledge of the spatial and temporal control of GI cell fates might ultimately contribute to developing cell-based therapies for treatment of these disorders.

Acknowledgments

The authors thank all the reviewers for their encouragement and in particular reviewers 1 and 3 for extremely constructive suggestions. V.A.M. thanks Peter M. Carson for enduring support.

References

1. 1Gilbert SF. Developmental biology. Sunderland, MA: Sinauer Associates, Inc; 2000;.

2. 2Yasugi S. Role of epithelial-mesenchymal interactions in differentiation of epithelium of vertebrate digestive organs. Dev Growth Differ. 1993;35:1–9.

3. 3Roberts DJ, Smith DM, Goff DJ, et al. Epithelial-mesenchymal signaling during the regionalization of the chick gut. Development. 1998;125:2791–2801. MEDLINE

4. 4Ekker SC, McGrew LL, Lai CJ, et al. Distinct expression and shared activities of members of the hedgehog gene family of Xenopus laevis. Development. 1995;121:2337–2347. MEDLINE

5. 5Roberts DJ, Johnson RL, Burke AC, et al. Sonic hedgehog is an endodermal signal inducing Bmp-4 and Hox genes during induction and regionalization of the chick hindgut. Development. 1995;121:3163–3174. MEDLINE

6. 6Bitgood MJ, McMahon AP. Hedgehog and Bmp genes are coexpressed at many diverse sites of cell-cell interaction in the mouse embryo. Dev Biol. 1995;172:126–138. MEDLINE | CrossRef

7. 7Echelard Y, Epstein DJ, St-Jacques B, et al. Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell. 1993;75:1417–1430. MEDLINE | CrossRef

8. 8Stolow MA, Shi YB. Xenopus sonic hedgehog as a potential morphogen during embryogenesis and thyroid hormone-dependent metamorphosis. Nucleic Acids Res. 1995;23:2555–2562. MEDLINE

9. 9Kedinger M, Simon-Assmann P, Lacroix B. [Development of the digestive function: regulation of the maturation of intestinal brush border enzymes]. Reprod Nutr Dev. 1986;26:691–702.

10. 10Apelqvist A, Ahlgren U, Edlund H. Sonic hedgehog directs specialised mesoderm differentiation in the intestine and pancreas. Curr Biol. 1997;7:801–804. MEDLINE | CrossRef

11. 11Winnier G, Blessing M, Labosky PA, et al. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 1995;9:2105–2116. MEDLINE | CrossRef

12. 12Wells JM, Melton DA. Vertebrate endoderm development. Annu Rev Cell Dev Biol. 1999;15:393–410. MEDLINE | CrossRef

13. 13Zaret KS. Regulatory phases of early liver development: paradigms of organogenesis. Nat Rev Genet. 2002;3:499–512. MEDLINE | CrossRef

14. 14McLin VA, Zorn AM. Molecular control of liver development. Clin Liver Dis. 2006;10:1–25v. Abstract | Full Text | Full-Text PDF (563 KB) | CrossRef

15. 15Cano DA, Hebrok M, Zenker M. Pancreatic development and disease. Gastroenterology. 2007;132:745–762. Full Text | Full-Text PDF (1113 KB) | CrossRef

16. 16Montgomery RK, Mulberg AE, Grand RJ. Development of the human gastrointestinal tract: twenty years of progress. Gastroenterology. 1999;116:702–731. Abstract | Full Text | Full-Text PDF (551 KB) | CrossRef

17. 17Joyce NC, Haire MF, Palade GE. Morphologic and biochemical evidence for a contractile cell network within the rat intestinal mucosa. Gastroenterology. 1987;92:68–81. Abstract | Full-Text PDF (3878 KB)

18. 18Sappino AP, Dietrich PY, Skalli O, et al. Colonic pericryptal fibroblasts (Differentiation pattern in embryogenesis and phenotypic modulation in epithelial proliferative lesions). Virchows Arch A Pathol Anat Histopathol. 1989;415:551–557. MEDLINE

19. 19Plateroti M, Rubin DC, Duluc I, et al. Subepithelial fibroblast cell lines from different levels of gut axis display regional characteristics. Am J Physiol. 1998;274:G945–G954. MEDLINE

20. 20Hall PA, Coates PJ, Ansari B, et al. Regulation of cell number in the mammalian gastrointestinal tract: the importance of apoptosis. J Cell Sci. 1994;107:3569–3577.

21. 21Wallace KN, Pack M. Unique and conserved aspects of gut development in zebrafish. Dev Biol. 2003;255:12–29. MEDLINE | CrossRef

22. 22Smith VV, Milla PJ. Histological phenotypes of enteric smooth muscle disease causing functional intestinal obstruction in childhood. Histopathology. 1997;31:112–122. MEDLINE | CrossRef

23. 23Hohmann B, Holzwarth M, Saur G, et al. Neonatal necrotizing enterocolitis and co-existing defect of the intestinal musculature. Eur J Pediatr. 1993;152:540–541. MEDLINE | CrossRef

24. 24Izraeli S, Freud E, Mor C, et al. Neonatal intestinal perforation due to congenital defects in the intestinal muscularis. Eur J Pediatr. 1992;151:300–303. MEDLINE | CrossRef

25. 25McHugh KM. Molecular analysis of smooth muscle development in the mouse. Dev Dyn. 1995;204:278–290. MEDLINE

26. 26McHugh KM. Molecular analysis of gastrointestinal smooth muscle development. J Pediatr Gastroenterol Nutr. 1996;23:379–394. MEDLINE | CrossRef

27. 27Wallace KN, Akhter S, Smith EM, et al. Intestinal growth and differentiation in zebrafish. Mech Dev. 2005;122:157–173. MEDLINE | CrossRef

28. 28Andrew A. The origin of intramural ganglia. II (The trunk neural crest as a source of enteric ganglion cells). J Anat. 1969;105:89–101. MEDLINE

29. 29Cochard P, Le Douarin NM. Development of the intrinsic innervation of the gut. Scand J Gastroenterol Suppl. 1982;71:1–14. MEDLINE

30. 30Gershon MD, Payette RF, Rothman TP. Development of the enteric nervous system. Fed Proc. 1983;42:1620–1625. MEDLINE

31. 31Rothman TP, Nilaver G, Gershon MD. Colonization of the developing murine enteric nervous system and subsequent phenotypic expression by the precursors of peptidergic neurons. J Comp Neurol. 1984;225:13–23. MEDLINE | CrossRef

32. 32Henning SJ, Rubin DC, Shulman RJ. Ontogeny of the intestinal mucosa. Raven Press; 1994;.

33. 33Wells JM, Melton DA. Early mouse endoderm is patterned by soluble factors from adjacent germ layers. Development. 2000;127:1563–1572. MEDLINE

34. 34Dessimoz J, Opoka R, Kordich JJ, et al. FGF signaling is necessary for establishing gut tube domains along the anterior-posterior axis in vivo. Mech Dev. 2006;123:42–55. MEDLINE | CrossRef

35. 35Sukegawa A, Narita T, Kameda T, et al. The concentric structure of the developing gut is regulated by Sonic hedgehog derived from endodermal epithelium. Development. 2000;127:1971–1980. MEDLINE

36. 36Tseng HT, Shah R, Jamrich M. Function and regulation of FoxF1 during Xenopus gut development. Development. 2004;131:3637–3647. MEDLINE | CrossRef

37. 37Zaffran S, Kuchler A, Lee HH, et al. biniou (FoxF), a central component in a regulatory network controlling visceral mesoderm development and midgut morphogenesis in Drosophila. Genes Dev. 2001;15:2900–2915. MEDLINE

38. 38Mahlapuu M, Enerback S, Carlsson P. Haploinsufficiency of the forkhead gene Foxf1, a target for sonic hedgehog signaling, causes lung and foregut malformations. Development. 2001;128:2397–2406. MEDLINE

39. 39Jakobsen JS, Braun M, Astorga J, et al. Temporal ChIP-on-chip reveals Biniou as a universal regulator of the visceral muscle transcriptional network. Genes Dev. 2007;21:2448–2460. CrossRef

40. 40Mahlapuu M, Pelto-Huikko M, Aitola M, et al. FREAC-1 contains a cell-type-specific transcriptional activation domain and is expressed in epithelial-mesenchymal interfaces. Dev Biol. 1998;202:183–195. MEDLINE | CrossRef

41. 41Mahlapuu M, Ormestad M, Enerback S, et al. The forkhead transcription factor Foxf1 is required for differentiation of extra-embryonic and lateral plate mesoderm. Development. 2001;128:155–166. MEDLINE

42. 42Astorga J, Carlsson P. Hedgehog induction of murine vasculogenesis is mediated by Foxf1 and Bmp4. Development. 2007;134:3753–3761. CrossRef

43. 43Kalinichenko VV, Zhou Y, Bhattacharyya D, et al. Haploinsufficiency of the mouse Forkhead Box f1 gene causes defects in gall bladder development. J Biol Chem. 2002;277:12369–12374. MEDLINE | CrossRef

44. 44Aitola M, Carlsson P, Mahlapuu M, et al. Forkhead transcription factor FoxF2 is expressed in mesodermal tissues involved in epithelio-mesenchymal interactions. Dev Dyn. 2000;218:136–149. MEDLINE | CrossRef

45. 45Ormestad M, Astorga J, Carlsson P. Differences in the embryonic expression patterns of mouse Foxf1 and -2 match their distinct mutant phenotypes. Dev Dyn. 2004;229:328–333. MEDLINE | CrossRef

46. 46Ormestad M, Astorga J, Landgren H, et al. Foxf1 and Foxf2 control murine gut development by limiting mesenchymal Wnt signaling and promoting extracellular matrix production. Development. 2006;133:833–843. MEDLINE | CrossRef

47. 47Sandmann T, Jensen LJ, Jakobsen JS, et al. A temporal map of transcription factor activity: mef2 directly regulates target genes at all stages of muscle development. Dev Cell. 2006;10:797–807. MEDLINE | CrossRef

48. 48Krauss S, Concordet JP, Ingham PW. A functionally conserved homolog of the Drosophila segment polarity gene hh is expressed in tissues with polarizing activity in zebrafish embryos. Cell. 1993;75:1431–1444. MEDLINE | CrossRef

49. 49Sakiyama J, Yokouchi Y, Kuroiwa A. HoxA and HoxB cluster genes subdivide the digestive tract into morphological domains during chick development. Mech Dev. 2001;101:233–236. MEDLINE | CrossRef

50. 50Sakiyama J, Yokouchi Y, Kuroiwa A. Coordinated expression of Hoxb genes and signaling molecules during development of the chick respiratory tract. Dev Biol. 2000;227:12–27. MEDLINE | CrossRef

51. 51Grapin-Botton A, Melton DA. Endoderm development: from patterning to organogenesis. Trends Genet. 2000;16:124–130. MEDLINE | CrossRef

52. 52Beck F, Tata F, Chawengsaksophak K. Homeobox genes and gut development. Bioessays. 2000;22:431–441. MEDLINE | CrossRef

53. 53Yokouchi Y, Sakiyama J, Kuroiwa A. Coordinated expression of Abd-B subfamily genes of the HoxA cluster in the developing digestive tract of chick embryo. Dev Biol. 1995;169:76–89. MEDLINE | CrossRef

54. 54Stanfel MN, Moses KA, Schwartz RJ, et al. Regulation of organ development by the NKX-homeodomain factors: an NKX code. Cell Mol Biol (Noisy-le-grand). 2005;(Suppl 51):OL785–OL799.

55. 55Park M, Lewis C, Turbay D, et al. Differential rescue of visceral and cardiac defects in Drosophila by vertebrate tinman-related genes. Proc Natl Acad Sci U S A. 1998;95:9366–9371. MEDLINE | CrossRef

56. 56Zakany J, Duboule D. Hox genes and the making of sphincters. Nature. 1999;401:761–762. MEDLINE | CrossRef

57. 57Zacchetti G, Duboule D, Zakany J. Hox gene function in vertebrate gut morphogenesis: the case of the caecum. Development. 2007;134:3967–3973. CrossRef

58. 58Warot X, Fromental-Ramain C, Fraulob V, et al. Gene dosage-dependent effects of the Hoxa-13 and Hoxd-13 mutations on morphogenesis of the terminal parts of the digestive and urogenital tracts. Development. 1997;124:4781–4791. MEDLINE

59. 59Goodman FR, Bacchelli C, Brady AF, et al. Novel HOXA13 mutations and the phenotypic spectrum of hand-foot-genital syndrome. Am J Hum Genet. 2000;67:197–202. MEDLINE | CrossRef

60. 60Goodman FR, Scambler PJ. Human HOX gene mutations. Clin Genet. 2001;59:1–11. MEDLINE | CrossRef

61. 61Tennyson VM, Gershon MD, Sherman DL, et al. Structural abnormalities associated with congenital megacolon in transgenic mice that overexpress the Hoxa-4 gene. Dev Dyn. 1993;198:28–53. MEDLINE

62. 62Aubin J, Dery U, Lemieux M, et al. Stomach regional specification requires Hoxa5-driven mesenchymal-epithelial signaling. Development. 2002;129:4075–4087. MEDLINE

63. 63Wang Z, Dolle P, Cardoso WV, et al. Retinoic acid regulates morphogenesis and patterning of posterior foregut derivatives. Dev Biol. 2006;297:433–445. MEDLINE | CrossRef

64. 64Kumar M, Jordan N, Melton D, et al. Signals from lateral plate mesoderm instruct endoderm toward a pancreatic fate. Dev Biol. 2003;259:109–122. MEDLINE | CrossRef

65. 65Ryan JF, Baxevanis AD. Hox, Wnt, and the evolution of the primary body axis: insights from the early-divergent phyla. Biol Direct. 2007;2:37.

66. 66Lohnes D. The Cdx1 homeodomain protein: an integrator of posterior signaling in the mouse. Bioessays. 2003;25:971–980. MEDLINE | CrossRef

67. 67Pilon N, Oh K, Sylvestre JR, et al. Cdx4 is a direct target of the canonical Wnt pathway. Dev Biol. 2006;289:55–63. MEDLINE | CrossRef

68. 68Pilon N, Oh K, Sylvestre JR, et al. Wnt signaling is a key mediator of Cdx1 expression in vivo. Development. 2007;134:2315–2323. MEDLINE | CrossRef

69. 69Deschamps J, van Nes J. Developmental regulation of the Hox genes during axial morphogenesis in the mouse. Development. 2005;132:2931–2942. MEDLINE | CrossRef

70. 70Mandhan P, Quan QB, Beasley S, et al. Sonic hedgehog, BMP4, and Hox genes in the development of anorectal malformations in Ethylenethiourea-exposed fetal rats. J Pediatr Surg. 2006;41:2041–2045. Abstract | Full Text | Full-Text PDF (635 KB) | CrossRef

71. 71Ramalho-Santos M, Melton DA, McMahon AP. Hedgehog signals regulate multiple aspects of gastrointestinal development. Development. 2000;127:2763–2772. MEDLINE

72. 72Madison BB, McKenna LB, Dolson D, et al. FoxF1 and FoxL1 link hedgehog signaling and the control of epithelial proliferation in the developing stomach and intestine. J Biol Chem. 2009;284:5936–5944. CrossRef

73. 73Iafolla AK, Kahler SG. Megalencephaly in the neonatal period as the initial manifestation of glutaric aciduria type I. J Pediatr. 1989;114:1004–1006. Full-Text PDF (735 KB) | CrossRef

74. 74Piekarski DH, Stephens FD. The association and embryogenesis of tracheo-oesophageal and anorectal anomalies. Prog Pediatr Surg. 1976;9:63–76. MEDLINE

75. 75Nielsen C, Murtaugh LC, Chyung JC, et al. Gizzard formation and the role of Bapx1. Dev Biol. 2001;231:164–174. MEDLINE | CrossRef

76. 76Yasugi S, Matsushita S, Mizuno T. Gland formation induced in the allantoic and small-intestinal endoderm by the proventricular mesenchyme is not coupled with pepsinogen expression. Differentiation. 1985;30:47–52. CrossRef

77. 77Kedinger M, Simon-Assmann P, Bouziges F, et al. Epithelial-mesenchymal interactions in intestinal epithelial differentiation. Scand J Gastroenterol Suppl. 1988;151:62–69. MEDLINE

78. 78Haffen K, Kedinger M, Simon-Assmann PM, et al. Mesenchyme-dependent differentiation of intestinal brush-border enzymes in the gizzard endoderm of the chick embryo. Prog Clin Biol Res. 1982;85:261–270.

79. 79Ishizuya-Oka A, Mizuno T. Intestinal cytodifferentiation in vitro of chick stomach endoderm induced by the duodenal mesenchyme. J Embryol Exp Morphol. 1984;82:163–176. MEDLINE

80. 80Roberts DJ. Molecular mechanisms of development of the gastrointestinal tract. Dev Dyn. 2000;219:109–120. MEDLINE

81. 81Goldstein 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. MEDLINE | CrossRef

82. 82Verzi MP, Stanfel MN, Moses KA, et al. Role of the homeodomain transcription factor Bapx1 in mouse distal stomach development. Gastroenterology. 2009 Jan 14;[Epub ahead of print].

83. 83Kim 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. MEDLINE | CrossRef

84. 84Moniot B, Biau S, Faure S, et al. SOX9 specifies the pyloric sphincter epithelium through mesenchymal-epithelial signals. Development. 2004;131:3795–3804. MEDLINE | CrossRef

85. 85Theodosiou NA, Tabin CJ. Sox9 and Nkx2.5 determine the pyloric sphincter epithelium under the control of BMP signaling. Dev Biol. 2005;279:481–490. MEDLINE | CrossRef

86. 86Smith DM, Nielsen C, Tabin CJ, et al. Roles of BMP signaling and Nkx2.5 in patterning at the chick midgut-foregut boundary. Development. 2000;127:3671–3681. MEDLINE

87. 87Henning SJ, Rubin DC, Shulman RJ. Ontogeny of the intestinal mucosa. In: LR J editors. Physiology of the gastrointestinal tract. New York, NY: Raven Press; 1994;p. 571–610.

88. 88Lacroix B, Kedinger M, Simon-Assmann PM, et al. Effects of human fetal gastroenteric mesenchymal cells on some developmental aspects of animal gut endoderm. Differentiation. 1984;28:129–135. MEDLINE | CrossRef

89. 89Ratineau C, Duluc I, Pourreyron C, et al. Endoderm- and mesenchyme-dependent commitment of the differentiated epithelial cell types in the developing intestine of rat. Differentiation. 2003;71:163–169. MEDLINE | CrossRef

90. 90Hayashi K, Yasugi S, Mizuno T. Pepsinogen gene transcription induced in heterologous epithelial-mesenchymal recombinations of chicken endoderms and glandular stomach mesenchyme. Development. 1988;103:725–731. MEDLINE

91. 91Rankin CT, Bunton T, Lawler AM, et al. Regulation of left-right patterning in mice by growth/differentiation factor-1. Nat Genet. 2000;24:262–265. MEDLINE | CrossRef

92. 92Collignon J, Varlet I, Robertson EJ. Relationship between asymmetric nodal expression and the direction of embryonic turning. Nature. 1996;381:155–158. MEDLINE | CrossRef

93. 93Sampath K, Cheng AM, Frisch A, et al. Functional differences among Xenopus nodal-related genes in left-right axis determination. Development. 1997;124:3293–3302. MEDLINE

94. 94Yashiro K, Saijoh Y, Sakuma R, et al. Distinct transcriptional regulation and phylogenetic divergence of human LEFTY genes. Genes Cells. 2000;5:343–357. MEDLINE | CrossRef

95. 95Logan M, Pagan-Westphal SM, Smith DM, et al. The transcription factor Pitx2 mediates situs-specific morphogenesis in response to left-right asymmetric signals. Cell. 1998;94:307–317. MEDLINE | CrossRef

96. 96Yoshioka H, Meno C, Koshiba K, et al. Pitx2, a bicoid-type homeobox gene, is involved in a lefty-signaling pathway in determination of left-right asymmetry. Cell. 1998;94:299–305. MEDLINE | CrossRef

97. 97Davis NM, Kurpios NA, Sun X, et al. The chirality of gut rotation derives from left-right asymmetric changes in the architecture of the dorsal mesentery. Dev Cell. 2008;15:134–145. CrossRef

98. 98Lipscomb K, Schmitt C, Sablyak A, et al. Role for retinoid signaling in left-right asymmetric digestive organ morphogenesis. Dev Dyn. 2006;235:2266–2275. MEDLINE | CrossRef

99. 99El-Mounayri O, Triplett JW, Yates CW, et al. Regulation of smooth muscle-specific gene expression by homeodomain proteins, Hoxa10 and Hoxb8. J Biol Chem. 2005;280:25854–25863. MEDLINE | CrossRef

100. 100Louvard D, Kedinger M, Hauri HP. The differentiating intestinal epithelial cell: establishment and maintenance of functions through interactions between cellular structures. Annu Rev Cell Biol. 1992;8:157–195. MEDLINE

101. 101Kedinger M, Simon-Assmann PM, Lacroix B, et al. Fetal gut mesenchyme induces differentiation of cultured intestinal endodermal and crypt cells. Dev Biol. 1986;113:474–483. MEDLINE | CrossRef

102. 102Wang LC, Nassir F, Liu ZY, et al. Disruption of hedgehog signaling reveals a novel role in intestinal morphogenesis and intestinal-specific lipid metabolism in mice. Gastroenterology. 2002;122:469–482. Abstract | Full Text | Full-Text PDF (1143 KB) | CrossRef

103. 103Madison BB, Braunstein K, Kuizon E, et al. Epithelial hedgehog signals pattern the intestinal crypt-villus axis. Development. 2005;132:279–289. MEDLINE | CrossRef

104. 104Mishina Y. Function of bone morphogenetic protein signaling during mouse development. Front Biosci. 2003;8:d855–d869. CrossRef

105. 105Faure S, de Santa Barbara P, Roberts DJ, et al. Endogenous patterns of BMP signaling during early chick development. Dev Biol. 2002;244:44–65. MEDLINE | CrossRef

106. 106De Santa Barbara P, Williams J, Goldstein AM, et al. Bone morphogenetic protein signaling pathway plays multiple roles during gastrointestinal tract development. Dev Dyn. 2005;234:312–322. MEDLINE | CrossRef

107. 107Batts LE, Polk DB, Dubois RN, et al. Bmp signaling is required for intestinal growth and morphogenesis. Dev Dyn. 2006;235:1563–1570. MEDLINE | CrossRef

108. 108He XC, Zhang J, Tong WG, et al. BMP signaling inhibits intestinal stem cell self-renewal through suppression of Wnt-beta-catenin signaling. Nat Genet. 2004;36:1117–1121. MEDLINE | CrossRef

109. 109Auclair BA, Benoit YD, Rivard N, et al. Bone morphogenetic protein signaling is essential for terminal differentiation of the intestinal secretory cell lineage. Gastroenterology. 2007;133:887–896. Abstract | Full Text | Full-Text PDF (3695 KB) | CrossRef

110. 110Haramis 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. CrossRef

111. 111Ishizuya-Oka A, Hasebe T, Shimizu K, et al. Shh/BMP-4 signaling pathway is essential for intestinal epithelial development during Xenopus larval-to-adult remodeling. Dev Dyn. 2006;235:3240–3249. MEDLINE | CrossRef

112. 112Ishizuya-Oka A, Ueda S, Amano T, et al. Thyroid-hormone-dependent and fibroblast-specific expression of BMP-4 correlates with adult epithelial development during amphibian intestinal remodeling. Cell Tissue Res. 2001;303:187–195. CrossRef

113. 113Pinto D, Clevers H. Wnt control of stem cells and differentiation in the intestinal epithelium. Exp Cell Res. 2005;306:357–363. MEDLINE | CrossRef

114. 114Pinto D, Clevers H. Wnt, stem cells and cancer in the intestine. Biol Cell. 2005;97:185–196. MEDLINE | CrossRef

115. 115Pinto D, Gregorieff A, Begthel H, et al. Canonical Wnt signals are essential for homeostasis of the intestinal epithelium. Genes Dev. 2003;17:1709–1713. MEDLINE | CrossRef

116. 116Clevers H. Wnt/beta-catenin signaling in development and disease. Cell. 2006;127:469–480. MEDLINE | CrossRef

117. 117van Es JH, Jay P, Gregorieff A, et al. Wnt signalling induces maturation of Paneth cells in intestinal crypts. Nat Cell Biol. 2005;7:381–386. MEDLINE | CrossRef

118. 118Andoh A, Bamba S, Brittan M, et al. Role of intestinal subepithelial myofibroblasts in inflammation and regenerative response in the gut. Pharmacol Ther. 2007;114:94–106. MEDLINE | CrossRef

119. 119Gregorieff A, Pinto D, Begthel H, et al. Expression pattern of Wnt signaling components in the adult intestine. Gastroenterology. 2005;129:626–638. Abstract | Full Text | Full-Text PDF (1071 KB) | CrossRef

120. 120Theodosiou NA, Tabin CJ. Wnt signaling during development of the gastrointestinal tract. Dev Biol. 2003;259:258–271. MEDLINE | CrossRef

121. 121Muncan V, Faro A, Haramis AP, et al. T-cell factor 4 (Tcf7l2) maintains proliferative compartments in zebrafish intestine. EMBO Rep. 2007;8:966–973. CrossRef

122. 122Barker N, Huls G, Korinek V, et al. Restricted high level expression of Tcf-4 protein in intestinal and mammary gland epithelium. Am J Pathol. 1999;154:29–35. MEDLINE

123. 123Booth C, Brady G, Potten CS. Crowd control in the crypt. Nat Med. 2002;8:1360–1361. MEDLINE | CrossRef

124. 124Kim BM, Mao J, Taketo MM, et al. Phases of canonical Wnt signaling during the development of mouse intestinal epithelium. Gastroenterology. 2007;133:529–538. Abstract | Full Text | Full-Text PDF (2534 KB) | CrossRef

125. 125Cervantes S, Yamaguchi TP, Hebrok M. Wnt5a is essential for intestinal elongation in mice. Dev Biol. 2009;326:285–294. CrossRef

126. 126Geske MJ, Zhang X, Patel KK, et al. Fgf9 signaling regulates small intestinal elongation and mesenchymal development. Development. 2008;135:2959–2968. CrossRef

127. 127Kaestner KH, Silberg DG, Traber PG, et al. The mesenchymal winged helix transcription factor Fkh6 is required for the control of gastrointestinal proliferation and differentiation. Genes Dev. 1997;11:1583–1595. MEDLINE | CrossRef

128. 128Perreault N, Sackett SD, Katz JP, et al. Foxl1 is a mesenchymal Modifier of Min in carcinogenesis of stomach and colon. Genes Dev. 2005;19:311–315. MEDLINE | CrossRef

129. 129Katz JP, Perreault N, Goldstein BG, et al. Foxl1 null mice have abnormal intestinal epithelia, postnatal growth retardation, and defective intestinal glucose uptake. Am J Physiol Gastrointest Liver Physiol. 2004;287:G856–G864. MEDLINE | CrossRef

130. 130Oka T, Shiojima I, Monzen K, et al. Fibroblast growth factor plays a critical role in SM22alpha expression during Xenopus embryogenesis. Arterioscler Thromb Vasc Biol. 2000;20:907–914. MEDLINE

131. 131Saint-Jeannet JP, Thiery JP, Koteliansky VE. Effect of an inhibitory mutant of the FGF receptor on mesoderm-derived alpha-smooth muscle actin-expressing cells in Xenopus embryo. Dev Biol. 1994;164:374–382. MEDLINE | CrossRef

132. 132Li X, Madison BB, Zacharias W, et al. Deconvoluting the intestine: molecular evidence for a major role of the mesenchyme in the modulation of signaling cross talk. Physiol Genomics. 2007;29:290–301. CrossRef

133. 133Vidrich A, Buzan JM, Ilo C, et al. Fibroblast growth factor receptor-3 is expressed in undifferentiated intestinal epithelial cells during murine crypt morphogenesis. Dev Dyn. 2004;230:114–123. MEDLINE | CrossRef

134. 134Zhang X, Stappenbeck TS, White AC, et al. Reciprocal epithelial-mesenchymal FGF signaling is required for cecal development. Development. 2006;133:173–180. MEDLINE | CrossRef

135. 135Shulman RJ, Schanler RJ, Lau C, et al. Early feeding, antenatal glucocorticoids, and human milk decrease intestinal permeability in preterm infants. Pediatr Res. 1998;44:519–523. MEDLINE

136. 136Stallmach A, Hahn U, Merker HJ, et al. Differentiation of rat intestinal epithelial cells is induced by organotypic mesenchymal cells in vitro. Gut. 1989;30:959–970. MEDLINE | CrossRef

137. 137Kedinger M, Simon-Assmann P, Alexandre E, et al. Importance of a fibroblastic support for in vitro differentiation of intestinal endodermal cells and for their response to glucocorticoids. Cell Differ. 1987;20:171–182. MEDLINE | CrossRef

138. 138Foltzer-Jourdainne C, Kedinger M, Raul F. Perinatal expression of brush-border hydrolases in rat colon: hormonal and tissue regulations. Am J Physiol. 1989;257:G496–G503. MEDLINE

139. 139Walsh MJ, LeLeiko NS, Sterling KM. Regulation of types I, III, and IV procollagen mRNA synthesis in glucocorticoid-mediated intestinal development. J Biol Chem. 1987;262:10814–10818. MEDLINE

140. 140Simon-Assmann P, Kedinger M, De Arcangelis A, et al. Extracellular matrix components in intestinal development. Experientia. 1995;51:883–900. MEDLINE | CrossRef

141. 141Weiser MM, Sykes DE, Killen PD. Rat intestinal basement membrane synthesis: Epithelial versus nonepithelial contributions. Lab Invest. 1990;62:325–330. MEDLINE

142. 142Simo P, Simon-Assmann P, Arnold C, et al. Mesenchyme-mediated effect of dexamethasone on laminin in cocultures of embryonic gut epithelial cells and mesenchyme-derived cells. J Cell Sci. 1992;101:161–171.

143. 143Solomon NS, Gartner H, Oesterreicher TJ, Henning SJ. Development of glucocorticoid-responsiveness in mouse intestine. Pediatr Res. 2001;49:782–788. MEDLINE

144. 144Gartner H, Graul MC, Oesterreicher TJ, et al. Development of the fetal intestine in mice lacking the glucocorticoid receptor (GR). J Cell Physiol. 2003;194:80–87. MEDLINE | CrossRef

145. 145Fraichard A, Chassande O, Plateroti M, et al. The T3R alpha gene encoding a thyroid hormone receptor is essential for post-natal development and thyroid hormone production. EMBO J. 1997;16:4412–4420. MEDLINE | CrossRef

146. 146Schreiber AM, Cai L, Brown DD. Remodeling of the intestine during metamorphosis of Xenopus laevis. Proc Natl Acad Sci U S A. 2005;102:3720–3725. MEDLINE | CrossRef

147. 147Leeper LL, McDonald MC, Heath JP, et al. Sucrase-isomaltase ontogeny: synergism between glucocorticoids and thyroxine reflects increased mRNA and no change in cell migration. Biochem Biophys Res Commun. 1998;246:765–770. CrossRef

148. 148McDonald MC, Henning SJ. Synergistic effects of thyroxine and dexamethasone on enzyme ontogeny in rat small intestine. Pediatr Res. 1992;32:306–311. MEDLINE

149. 149Su Y, Shi Y, Shi YB. Cyclosporin A but not FK506 inhibits thyroid hormone-induced apoptosis in tadpole intestinal epithelium. FASEB J. 1997;11:559–565. MEDLINE

150. 150Su Y, Shi Y, Stolow MA, et al. Thyroid hormone induces apoptosis in primary cell cultures of tadpole intestine: cell type specificity and effects of extracellular matrix. J Cell Biol. 1997;139:1533–1543. MEDLINE | CrossRef

151. 151Ishizuya-Oka A, Shi YB. Regulation of adult intestinal epithelial stem cell development by thyroid hormone during Xenopus laevis metamorphosis. Dev Dyn. 2007;236:3358–3368. CrossRef

152. 152Quasnichka H, Slater SC, Beeching CA, et al. Regulation of smooth muscle cell proliferation by beta-catenin/T-cell factor signaling involves modulation of cyclin D1 and p21 expression. Circ Res. 2006;99:1329–1337. CrossRef

153. 153Karlsson L, Lindahl P, Heath JK, et al. Abnormal gastrointestinal development in PDGF-A and PDGFR-(alpha) deficient mice implicates a novel mesenchymal structure with putative instructive properties in villus morphogenesis. Development. 2000;127:3457–3466. MEDLINE

154. 154Hirota S, Isozaki K, Moriyama Y, et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science. 1998;279:577–580. MEDLINE | CrossRef

155. 155Taniguchi M, Nishida T, Hirota S, et al. Effect of c-kit mutation on prognosis of gastrointestinal stromal tumors. Cancer Res. 1999;59:4297–4300. MEDLINE

156. 156Rubin BP. Gastrointestinal stromal tumours: an update. Histopathology. 2006;48:83–96. MEDLINE | CrossRef

157. 157Heinrich MC, Rubin BP, Longley BJ, et al. Biology and genetic aspects of gastrointestinal stromal tumors: KIT activation and cytogenetic alterations. Hum Pathol. 2002;33:484–495. Abstract | Full Text | Full-Text PDF (410 KB) | CrossRef

158. 158Heinrich MC, Corless CL, Duensing A, et al. PDGFRA activating mutations in gastrointestinal stromal tumors. Science. 2003;299:708–710. CrossRef

159. 159MacDonald RS. The role of insulin-like growth factors in small intestinal cell growth and development. Horm Metab Res. 1999;31:103–113. MEDLINE | CrossRef

160. 160Michaylira CZ, Simmons JG, Ramocki NM, et al. Suppressor of cytokine signaling-2 limits intestinal growth and enterotrophic actions of IGF-I in vivo. Am J Physiol Gastrointest Liver Physiol. 2006;291:G472–G481. MEDLINE | CrossRef

161. 161Fukamachi H, Narita T, Yahagi N, et al. Endothelin-3 controls growth of colonic epithelial cells by mediating epithelial-mesenchymal interaction. Dev Growth Differ. 2005;47:573–580. MEDLINE | CrossRef

162. 162Wang Y, Wang L, Iordanov H, et al. Epimorphin(−/−) mice have increased intestinal growth, decreased susceptibility to dextran sodium sulfate colitis, and impaired spermatogenesis. J Clin Invest. 2006;116:1535–1546. MEDLINE | CrossRef

163. 163Hirai Y, Takebe K, Takashina M, et al. Epimorphin: a mesenchymal protein essential for epithelial morphogenesis. Cell. 1992;69:471–481. MEDLINE | CrossRef

164. 164Wallace KN, Dolan AC, Seiler C, et al. Mutation of smooth muscle myosin causes epithelial invasion and cystic expansion of the zebrafish intestine. Dev Cell. 2005;8:717–726. MEDLINE | CrossRef

165. 165Timpl R, Brown JC. Supramolecular assembly of basement membranes. Bioessays. 1996;18:123–132. MEDLINE | CrossRef

166. 166Shi YB, Fu L, Hasebe T, et al. Regulation of extracellular matrix remodeling and cell fate determination by matrix metalloproteinase stromelysin-3 during thyroid hormone-dependent post-embryonic development. Pharmacol Ther. 2007;116:391–400. CrossRef

167. 167El-Hodiri H, Bhatia-Dey N, Kenyon K, et al. Fox (forkhead) genes are involved in the dorso-ventral patterning of the Xenopus mesoderm. Int J Dev Biol. 2001;45:265–271. MEDLINE

168. 168Frasch M. Induction of visceral and cardiac mesoderm by ectodermal Dpp in the early Drosophila embryo. Nature. 1995;374:464–467. MEDLINE | CrossRef

169. 169Lee HH, Frasch M. Nuclear integration of positive Dpp signals, antagonistic Wg inputs and mesodermal competence factors during Drosophila visceral mesoderm induction. Development. 2005;132:1429–1442. MEDLINE | CrossRef

170. 170Stickney HL, Imai Y, Draper B, et al. Zebrafish bmp4 functions during late gastrulation to specify ventroposterior cell fates. Dev Biol. 2007;310:71–84. CrossRef

171. 171McMahon JA, Takada S, Zimmerman LB, et al. Noggin-mediated antagonism of BMP signaling is required for growth and patterning of the neural tube and somite. Genes Dev. 1998;12:1438–1452. MEDLINE | CrossRef

172. 172Monsoro-Burq AH, Duprez D, Watanabe Y, et al. The role of bone morphogenetic proteins in vertebral development. Development. 1996;122:3607–3616. MEDLINE

173. 173Xue L, Chen X, Chang Y, et al. Regulatory elements of the EKLF gene that direct erythroid cell-specific expression during mammalian development. Blood. 2004;103:4078–4083. MEDLINE | CrossRef

174. 174Davidson AJ, Zon LI. Turning mesoderm into blood: the formation of hematopoietic stem cells during embryogenesis. Curr Top Dev Biol. 2000;50:45–60. MEDLINE | CrossRef

175. 175Dosch R, Gawantka V, Delius H, et al. Bmp-4 acts as a morphogen in dorsoventral mesoderm patterning in Xenopus. Development. 1997;124:2325–2334. MEDLINE

176. 176Onichtchouk D, Glinka A, Niehrs C. Requirement for Xvent-1 and Xvent-2 gene function in dorsoventral patterning of Xenopus mesoderm. Development. 1998;125:1447–1456. MEDLINE

177. 177Ault KT, Dirksen ML, Jamrich M. A novel homeobox gene PV.1 mediates induction of ventral mesoderm in Xenopus embryos. Proc Natl Acad Sci U S A. 1996;93:6415–6420. MEDLINE | CrossRef

178. 178Onichtchouk D, Gawantka V, Dosch R, et al. The Xvent-2 homeobox gene is part of the BMP-4 signalling pathway controlling [correction of controling] dorsoventral patterning of Xenopus mesoderm. Development. 1996;122:3045–3053. MEDLINE

179. 179Kozmik Z, Holland LZ, Schubert M, et al. Characterization of Amphioxus AmphiVent, an evolutionarily conserved marker for chordate ventral mesoderm. Genesis. 2001;29:172–179. CrossRef

180. 180Bienz M, Tremml G. Domain of Ultrabithorax expression in Drosophila visceral mesoderm from autoregulation and exclusion. Nature. 1988;333:576–578. MEDLINE | CrossRef

181. 181Jagla K, Bellard M, Frasch M. A cluster of Drosophila homeobox genes involved in mesoderm differentiation programs. Bioessays. 2001;23:125–133. MEDLINE | CrossRef

182. 182Azpiazu N, Frasch M. tinman and bagpipe: two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 1993;7:1325–1340. MEDLINE | CrossRef

183. 183Azpiazu N, Lawrence PA, Vincent JP, et al. Segmentation and specification of the Drosophila mesoderm. Genes Dev. 1996;10:3183–3194. MEDLINE | CrossRef

184. 184Lo PC, Frasch M. bagpipe-Dependent expression of vimar, a novel Armadillo-repeats gene, in Drosophila visceral mesoderm. Mech Dev. 1998;72:65–75. MEDLINE | CrossRef

185. 185Newman CS, Grow MW, Cleaver O, et al. Xbap, a vertebrate gene related to bagpipe, is expressed in developing craniofacial structures and in anterior gut muscle. Dev Biol. 1997;181:223–233. MEDLINE | CrossRef

186. 186Newman CS, Krieg PA. The Xenopus bagpipe-related homeobox gene zampogna is expressed in the pharyngeal endoderm and the visceral musculature of the midgut. Dev Genes Evol. 1999;209:132–134. CrossRef

187. 187Nieuwkoop D, Faber J. Normal table of Xenopus laevis (Daudin). New York, NY: Garland Publishing Inc; 1994;.

188. 188Horb ME. Patterning the endoderm: the importance of neighbours. Bioessays. 2000;22:599–602. MEDLINE | CrossRef

189. 189Horb ME, Slack JM. Endoderm specification and differentiation in Xenopus embryos. Dev Biol. 2001;236:330–343. MEDLINE | CrossRef

190. 190Swindell EC, Eichele G. Retinoid metabolizing enzymes in development. Biofactors. 1999;10:85–89. MEDLINE | CrossRef

191. 191McLin VA, Rankin SA, Zorn AM. Repression of Wnt/beta-catenin signaling in the anterior endoderm is essential for liver and pancreas development. Development. 2007;134:2207–2217. MEDLINE | CrossRef

192. 192Glinka A, Wu W, Onichtchouk D, et al. Head induction by simultaneous repression of Bmp and Wnt signalling in Xenopus. Nature. 1997;389:517–519. MEDLINE | CrossRef

193. 193Okubo T, Hogan BL. Hyperactive Wnt signaling changes the developmental potential of embryonic lung endoderm. J Biol. 2004;3:11.

194. 194Crosnier C, Stamataki D, Lewis J. Organizing cell renewal in the intestine: stem cells, signals and combinatorial control. Nat Rev Genet. 2006;7:349–359. MEDLINE | CrossRef

195. 195Lints TJ, Hartley L, Parsons LM, et al. Mesoderm-specific expression of the divergent homeobox gene Hlx during murine embryogenesis. Dev Dyn. 1996;205:457–470. MEDLINE | CrossRef

196. 196Bates MD, Schatzman LC, Lints T, et al. Structural and functional characterization of the mouse Hlx homeobox gene. Mamm Genome. 2000;11:836–842. MEDLINE | CrossRef

197. 197Hentsch B, Lyons I, Li R, et al. Hlx homeo box gene is essential for an inductive tissue interaction that drives expansion of embryonic liver and gut. Genes Dev. 1996;10:70–79. MEDLINE | CrossRef

198. 198Pabst O, Schneider A, Brand T, et al. The mouse Nkx2-3 homeodomain gene is expressed in gut mesenchyme during pre- and postnatal mouse development. Dev Dyn. 1997;209:29–35. MEDLINE | CrossRef

199. 199Buchberger A, Pabst O, Brand T, et al. Chick NKx-2.3 represents a novel family member of vertebrate homologues to the Drosophila homeobox gene tinman: differential expression of cNKx-2.3 and cNKx-2.5 during heart and gut development. Mech Dev. 1996;56:151–163. MEDLINE | CrossRef

⁎ Department of Pediatrics, Section of Gastroenterology, Hepatology and Nutrition, Baylor College of Medicine, Houston, Texas

§ Department of Cellular and Molecular Biology, Baylor College of Medicine, Houston, Texas

∥ Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas

‡ Departments of Medicine and Cell and Molecular Physiology, University of North Carolina Chapel Hill, Chapel Hill, North Carolina

Reprint requests Address requests for reprints to: Valérie A. McLin, MD, Unite de Gastroenterologie, Hepatologie et Nutrition Pediatrique, Hopital des Enfants, CH-1211 Geneva, Switzerland

Conflicts of interest The authors disclose no conflicts.

Funding V.A.M. is supported by the National Institutes of Health (K08DK078656) and a Young Investigator Award from the Children's Digestive Health and Nutrition Foundation, S.J.H. is supported by the National Institutes of Health (R01DK069585), and M.J. is supported by the Retinal Research Foundation.

PII: S0016-5085(09)00370-9

doi:10.1053/j.gastro.2009.03.001

21 of 81