Drs. Chan and Chen contributed equally to this work.
Hoxb3 vagal neural crest-specific enhancer element for controlling enteric nervous system development
Article first published online: 14 MAR 2005
Copyright © 2005 Wiley-Liss, Inc.
Special Issue: Special Focus on Limb Development
Volume 233, Issue 2, pages 473–483, June 2005
How to Cite
Chan, K. K., Chen, Y. S., Yau, T. O., Fu, M., Lui, V. C. H., Tam, P. K. H. and Sham, M. H. (2005), Hoxb3 vagal neural crest-specific enhancer element for controlling enteric nervous system development. Dev. Dyn., 233: 473–483. doi: 10.1002/dvdy.20347
- Issue published online: 12 MAY 2005
- Article first published online: 14 MAR 2005
- Manuscript Accepted: 14 DEC 2004
- Manuscript Revised: 13 DEC 2004
- Manuscript Received: 15 JUL 2004
- Research Grants Council of Hong Kong SAR, China. Grant Number: HKU 7184/99M
- vagal neural crest;
- enteric nervous system;
- intestinal aganglionosis;
- enhancer element;
- transgenic mice
The neural and glial cells of the intrinsic ganglia of the enteric nervous system (ENS) are derived from the hindbrain neural crest at the vagal level. The Hoxb3 gene is expressed in the vagal neural crest and in the enteric ganglia of the developing gut during embryogenesis. We have identified a cis-acting enhancer element b3IIIa in the Hoxb3 gene locus. In this study, by transgenic mice analysis, we examined the tissue specificity of the b3IIIa enhancer element using the lacZ reporter gene, with emphasis on the vagal neural crest cells and their derivatives in the developing gut. We found that the b3IIIa-lacZ transgene marks only the vagal region and not the trunk or sacral region. Using cellular markers, we showed that the b3IIIa-lacZ transgene was expressed in a subset of enteric neuroblasts during early development of the gut, and the expression was maintained in differentiated neurons of the myenteric plexus at later stages. The specificity of the b3IIIa enhancer in directing gene expression in the developing ENS was further supported by genetic analysis using the Dom mutant, a spontaneous mouse model of Hirschsprung's disease characterized by the absence of enteric ganglia in the distal gut. The colonization of lacZ-expressing cells in the large intestine was incomplete in all the Dom/b3IIIa-lacZ hybrid mutants we examined. To our knowledge, this is the only vagal neural crest-specific genetic regulatory element identified to date. This element could be used for a variety of genetic manipulations and in establishing transgenic mouse models for studying the development of the ENS. Developmental Dynamics 233:473–483, 2005. © 2005 Wiley-Liss, Inc.
During mouse embryogenesis, the gut is formed as an endodermal tube surrounded by visceral mesoderm. The endodermal layer differentiates into epithelium with region-specific biochemical functions, while the visceral mesoderm differentiates to form the lamina propria, submucosa, and muscular wall of the intestine. At around 9.5 days post coitum (dpc), the neural crest cells from the vagal region start to migrate into the foregut. As the mouse embryo develops, these neural crest cells further migrate and proliferate in the gut and form the enteric nervous system (ENS) that innervates the musculature of the gastrointestinal tract. Genes of the Hox clusters are known to be expressed in both the endodermal and mesodermal components of the developing gut, but involvement of Hox genes in the vagal neural crest and ENS is less well understood.
During embryogenesis, Hox genes are expressed in a colinear manner along the length of the lateral plate mesoderm, which contributes to the mesodermal component of the gut (Beck, 2002; Yokouchi et al., 1995). Systematic studies of the expression of Hox genes showed that Hox genes in paralogous groups 4–9 were expressed in the mid- to posterior mesodermal domains of the gut, and the expression patterns correlated with the morphological subdivisions of the gut along the anteroposterior axis (Sekimoto et al., 1998; Pitera et al., 1999). Based on the colinear expression patterns, it was suggested that Hox genes constitute specific enteric Hox codes for patterning and specifying morphogenesis for the development of the gastrointestinal tract, in a manner that is conceptually similar to that in the axial mesoderm and hindbrain ectoderm. In a gene targeting experiment in which several HoxD genes (Hoxd4, d8, d9, d10, d11, d12, and d13) were deleted, the mutant mice lacked the ileocecal sphincter, suggesting that these HoxD genes are normally collectively required for the specification of the ileocecal sphincter structure (Zakany and Duboule, 1999). Other gene knockout and transgenic mice studies also demonstrated the functions of Hox genes in the specification of structures along the gastrointestinal tract (Wolgemuth et al., 1989; Warot et al., 1997).
In addition to their expression in mesoderm, Hox genes were shown to be expressed in the gut endoderm. The Hoxb1 gene is expressed in the foregut endoderm at 9.5 dpc, and the expression of Hoxb1 in the gut is regulated by a retinoic acid-dependent enhancer element (Huang et al., 1998). Several Hox genes, including Hoxa2, a3, b4, c5, c6, b8, b9, c9, a13, and d13, are also expressed in the gut endoderm in a region-specific manner (Grapin-Botton and Melton, 2000). However, the examination of Hox genes expression in the gut endoderm during embryogenesis has not been exhaustive. Based on the expression patterns known so far, there is no apparent regional colinearity between members of the Hox gene clusters and their expression patterns in the gut endoderm. It has been suggested that regional specification of the gastrointestinal tract depends on interactions between the endoderm-derived epithelium and the gut mesoderm (Duluc et al., 1994, 1997), and these interactions might be mediated by signaling molecules such as Shh and Bmp4 (Roberts et al., 1995, 1998). In a Hoxa5 knockout mouse mutant, there was evidence that Hoxa5 expression in the gut mesoderm was important for the specification of overlying gut endoderm. Loss of Hoxa5 function in the mesoderm led to impaired physiological function of the intestine in the mutant mice (Aubin et al., 1999).
Hox genes of paralogous groups 3 to 5 are expressed in the hindbrain neurectoderm at rhombomeres 6 to 8, at the level where vagal neural crest cells originate. These Hox genes are potential candidates that may be involved in the development of the ENS (Kapur et al., 2004). In a transgenic mouse model, overexpression of Hoxa4 in the gut using a viral promoter resulted in hypoganglionosis in the terminal colon in neonatal and adult transgenic mice (Wolgemuth et al., 1989; Tennyson et al., 1993). The transgenic mice that overexpressed Hoxa4 displayed abnormalities in the smooth muscle as well as in the innervation of the terminal bowel. In this Hoxa4 mutant, multiple tissues in the gut are affected and the megacolon phenotype is not just due to abnormal intrinsic enteric ganglia that are derived from the neural crest cells (Tennyson et al., 1998). Our study of the expression of the HOXB5 gene in human embryos showed that there are two separate and discontinuous mesenchymal expression domains of HOXB5 expression, which include a distal domain preceding the migratory neural crest cells, and a proximal domain that overlaps with the neural crest cells. When the neural crest cells start to differentiate, HOXB5 expression in the mesenchyme is switched off, but is activated in the neural crest-derived enteric ganglia where its expression is maintained until adult (Fu et al., 2003). Our detailed analysis of the dynamic patterns of HOXB5 expression suggests that Hox genes may play multiple roles in the developing gut. So far, there are only limited examples showing that Hox genes can affect ENS development. Whether Hox genes can control vagal neural crest and ENS development directly, or they act on the mesodermal and endodermal tissues in the gut and affect ENS development through microenvironmental cues, requires further elucidation.
In our transgenic mouse analysis to investigate the cis-acting components that regulate the dynamic patterns of Hoxb3 expression in mouse embryos, we identified a specific enhancer element IIIa in the Hoxb3 gene locus that directs the expression of the reporter gene in the anterior spinal cord and hindbrain up to rhombomere 6 (r6) and the associated vagal neural crest (Kwan et al., 2001; Yau et al., 2002). In this study, we examined the tissue specificity of the IIIa enhancer element (designated b3IIIa) using the lacZ reporter gene, focusing on the vagal neural crest cells and their derivatives in the gut. We also studied the expression of the b3IIIa-lacZ transgene in the Dom megacolon mutant mouse background (Lane and Liu, 1984; Herbarth et al., 1998; Southard-Smith et al., 1998) and provided evidence that the b3IIIa element has specific activities in the vagal neural crest and in the development of the ENS. The b3IIIa element, therefore, could be involved in normal regulation of Hoxb3 expression in the vagal neural crest and ENS. To our knowledge, this is the only vagal neural crest-specific genetic regulatory element identified to date, and this element could be used for a variety of genetic manipulations and in establishing transgenic mouse models for studying the development of the ENS.
Expression of Hoxb3 in the Developing Gut
We have shown previously that the Hoxb3 gene is expressed in the mesenchyme of the developing gut, with a high level of expression in the stomach of 12.5 dpc mouse embryo (Sham et al., 1992). To examine whether the Hoxb3 gene is expressed in the developing enteric nervous system, we performed RNA in situ hybridization on 11.5 and 13.5 dpc mouse embryo sections. Hoxb3 hybridization signals were detected in small clusters of cells in the mesenchyme at 11.5 dpc embryonic intestine (Fig. 1A). The Hoxb3 expression pattern was similar to that of Ret (Fig. 1C), which marked the neural crest cells and the enteric neurons. At 13.5 dpc, Ret expression was clearly detected in the neural crest-derived enteric ganglia in the myenteric region of the intestine (Fig. 1D). At this stage, the expression of Hoxb3 was also detected in the myenteric ganglia of the intestine (Fig. 1B). Therefore, in addition to the mesenchyme of the stomach, the Hoxb3 gene is expressed in the enteric neural crest in the developing intestine and Hoxb3 may also be involved in ENS development.
Expression of lacZ in b3IIIa-lacZ Transgenic Mice
In our previous study on the cis-regulation of Hoxb3, we have identified a 482-bp Sau3A fragment within element IIIa, which contained the critical region for element IIIa activity (Fig. 2A; Yau et al., 2002). We have established four independent b3IIIa-lacZ transgenic mouse lines using the 482-bp b3IIIa enhancer construct, which allowed us to examine the temporal and spatial expression patterns of the transgene in detail. All four transgenic mouse lines have the same expression patterns in the neural tube, with some differences in other tissues. In all the 10.5 dpc b3IIIa-lacZ transgenic embryos we examined, lacZ was expressed in the neural tube in the hindbrain up to rhombomere 6 (r6) and extending to the anterior spinal cord at the level of around somite 10; and in the associated neural crest which migrated ventrally to the mesenchyme of the third, fourth, and posterior branchial aches as well as the foregut (Figs. 2B–D, 3G). Of interest, transgenic embryos derived from line 2 also showed strong expression in the cardiac outflow tract (Fig. 2C). The cardiac expression was likely to be contributed by cardiac neural crest cells, which were derived from the neural tube at an axial level overlapping with the vagal neural crest. For transgenic embryos of line 3, the level of lacZ expression in r6 of the hindbrain was not up-regulated as in the other lines; there were also ectopic expression of lacZ in the posterior somites and hindlimb buds (Fig. 2D).
The temporal lacZ expression patterns in the neural tube and neural crest were the same in all four transgenic mouse lines. To illustrate the temporal and spatial expression patterns, we used whole-mount and sectioned X-gal–stained embryos derived from transgenic lines 1 and 4, which have identical expression patterns and represent the common patterns among the four mouse lines. In 8.5 dpc transgenic embryos, expression of the lacZ gene was restricted to the anterior neural tube, with an anterior limit at the developing hindbrain at the posterior rhombomeric sulcus (Fig. 3A). In 9.5 dpc embryos, lacZ was expressed in the hindbrain up to r6 and in the anterior spinal cord. At this stage, the associated neural crest, including the cardiac and vagal neural crest migrating to the third, fourth, and posterior branchial arches as well as the foregut, were positively stained with lacZ (Fig. 3B). Similar expression patterns were maintained in 10.5 dpc transgenic embryos, and the lacZ-marked vagal neural crest cells migrating to the foregut could be observed (Fig. 3C,G,J). By 12.5 dpc, lacZ expression in the neural tube was down-regulated (Fig. 3D) and no longer detectable in the neural tube of 13.5 and 14.5 dpc embryos (Fig. 3E,F). However, high levels of lacZ expression could be observed in the developing gut (Fig. 3D–L). We think that these lacZ-stained cells in the gut could be vagal neural crest-derived enteric ganglia; therefore, we further examined the lacZ expression profiles in the developing gastrointestinal tract of the transgenic embryos. Although the vagal neural crest also gives rise to part of the sympathetic nervous system (Durbec et al., 1996), we have not detected lacZ staining in the sympathetic ganglia of the b3IIIa-lacZ transgenic embryos.
Expression of lacZ in the Developing ENS
Intact gut specimens were dissected out from the b3IIIa-lacZ transgenic embryos of different stages, and the lacZ expression patterns were analyzed after whole-mount X-gal staining. At 10.5 dpc when the gut was a simple tube with a ventral loop in the midgut region, lacZ was already expressed in the foregut and proximal midgut (Fig. 4A). The lacZ-stained cells were sparsely distributed in the gut, resembling the distribution of enteric neural crest cells. At 11.5 dpc, the rudimentary esophagus and cecum were obvious and lacZ-stained cells were found in the esophagus and stomach; expression could be detected from the foregut down to the proximal part of the hindgut (Fig. 4B). At 12.5 dpc, the cecum could be easily recognized and the midgut was a convoluted tube. The lacZ-positive cells were sparse in the esophagus, but they could be readily observed in the stomach, small intestine, and approximately two thirds of the proximal large intestine (Fig. 4C). At 13.5 dpc, the developing gut had elongated, rotated to a great extent, and differentiated into regionally distinct segments. By this stage, only very few lacZ-expressing cells were present in the esophagus. The lacZ marked cells were mainly distributed from the stomach to nearly the distal end of the large intestine (Fig. 4D). At 14.5 dpc and 16.5 dpc, the gastrointestinal tract had further elongated and the distribution and density of transgene-expressing cells in the stomach and small intestine were similar to that at 13.5 dpc. From 14.5 dpc onward, the entire large intestine was fully colonized by the lacZ-marked cells. Furthermore, the intensity of lacZ staining was higher in the large intestine than other gut regions (Fig. 4E,F).
We have examined a large number of b3IIIa-lacZ transgenic embryos (10.5 dpc, n = 27; 11.5 dpc, n = 33; 12.5 dpc, n = 24; 13.5 dpc, n = 25; 14.5 dpc, n = 21; 16.5 dpc, n=30), and in summary, we found that the lacZ transgene expression patterns in the gastrointestinal tract of the transgenic embryos were consistent with the migration patterns of vagal neural crest cells that migrated to the gut, proliferated, and formed the enteric nervous system (Fig. 7A). Moreover, the intensity of lacZ staining in the marked cell population changed in different developmental stages. The intensity of lacZ staining was high in the stomach and small intestine at 11.5 dpc and 12.5 dpc, when the migration front of the lacZ-stained cells had reached the large intestine at these stages. Later, by 13.5 dpc, the intensity of lacZ staining in the stomach and small intestine was much reduced, but staining was stronger in the cecum and large intestine. By 14.5 dpc and 16.5 dpc, the highest lacZ staining intensity was found in the large intestine (Figs. 4, 7A).
Immunoreactivity of lacZ-Expressing Cells in the Gut
To examine the identity of the lacZ-expressing cells in the gastrointestinal tract of the b3IIIa-lacZ transgenic embryos, we performed a series of immunohistochemical staining on embryonic gut sections using antibodies specific for neural crest, neuronal, and glial cells of the enteric nervous system. All the four transgenic mouse lines showed similar lacZ staining and expression patterns of the cellular markers examined by immunostaining on gut sections. As shown in Figure 5A, PGP9.5, which is an early marker for neural crest and neuronal derivatives, was expressed in the myenteric plexus in 12.5 dpc embryonic gut. Most of the cells marked by PGP9.5 were also positively stained for lacZ activity. Similarly, when we examined 12.5 dpc and 14.5 dpc embryonic gut sections with p75NTR antibody, which marks undifferentiated enteric neural crest cells as well as enteric neurons, a ring of cells in the myenteric region were stained (Fig. 5B,C). At these stages, all the lacZ-stained cells had positive p75NTR immunoreactivity, suggesting that the transgene was expressed in the enteric neural crest and neuroblasts.
To examine transgene expression in differentiated cells of the enteric nervous system, we stained the embryonic gut sections with an antibody (2H3) against neurofilament NF-M, which marks nerve fibers of differentiated neurons. As shown in Figure 5E, in 12.5 dpc embryos, immunoreactivity for NF-M was observed in the enteric ganglia of the myenteric plexus, which were also positively stained for lacZ. The ring-like pattern formed by the lacZ-marked cells overlapped with the positive NF-M–stained nerve fibers. At 14.5dpc, the transgene-expressing cells were restricted to the myenteric plexus and there were more NF-M–stained nerve bundles (Fig. 5F). In 15.5 dpc and 16.5 dpc gut sections, the NF-M staining became more extensive as neuronal differentiation progressed and all the transgene-expressing cells were associated with NF-M–positive staining (Fig. 5G,H). Therefore, the b3IIIa-lacZ transgene coexpressed with neuronal markers of the developing enteric nervous system.
At 16.5 dpc, the glial cell marker glial fibrillary acidic protein (GFAP) started to express, glial cells in the myenteric plexus were distributed as a ring around the gut as shown by the positive GFAP immunoreactivity in embryonic gut sections (Fig. 5D). Except for a few of the glial cells that also expressed lacZ, most of the lacZ-positive cells appeared as clusters associated with the ring of glial cells (Fig. 5D). Therefore, expression of lacZ appeared to be switched off in differentiated glial cells of the ENS.
Pattern of Transgene Expression in the Dominant megacolon Hybrid Embryos
The Dom (Dominant megacolon) mutant mouse, which has a frame-shift mutation in the Sox10 locus, displays spotted pigmentation and enteric aganglionosis due to neural crest defects (Herbarth et al., 1998; Southard-Smith et al., 1998, 1999). We crossed the b3IIIa-lacZ transgenic mouse line with the Dom/+ mutant and generated hybrid mutant embryos heterozygous for both b3IIIa-lacZ and Dom loci and used the Dom mutant background to assess the specificity of the b3IIIa enhancer activity in the developing enteric nervous system. Analysis of the lacZ expression patterns in the gastrointestinal tract of these Dom/b3IIIa-lacZ hybrid mutants showed that the lacZ-stained cells were greatly reduced (Fig. 6). At 12.5 dpc when lacZ-expressing cells in b3IIIa-lacZ embryos had reached the middle of the large intestine (Fig. 4C), in Dom/b3IIIa-lacZ hybrid mutant guts (n = 7), the density of lacZ-positive cells were reduced along the gastrointestinal tract and the colonization of lacZ-positive cells were greatly retarded (Fig. 6A,B). The extent of retardation in the migration of lacZ-marked cells was different among different embryos, the migration front of the lacZ-stained cells varied between the small intestine and the cecum, but no lacZ-positive cells could be detected in the large intestine in the hybrid embryos examined (Fig. 6A,B). In 14.5 dpc hybrid mutant guts (n = 4), the distribution and intensity of lacZ-marked cells in the stomach and the small intestine were similar to those in the b3IIIa-lacZ embryonic guts, but a dramatic difference was observed in the cecum and hindgut region. In the Dom/b3IIIa-lacZ hybrid embryos, no lacZ-positive cells could be detected in the distal large intestine. Moreover, the distribution of lacZ-positive cells in the cecum was irregular, and in some cases, blank areas with regional absence of transgene-expressing cells were observed (Fig. 6C,D). In the hybrid mutant embryonic gut of 16.5dpc (n = 6), the density of lacZ-marked cells was significantly reduced. The limit of colonization of lacZ-marked cells along the length of the gut varied: in two of the hybrid embryos, no lacZ staining could be detected beyond the cecum; in the other four hybrid embryos, some lacZ-marked cells could be detected in the large intestine (Fig. 6E–G). However, no lacZ cells could be detected in the distal end of the large intestine in any of the hybrid embryos examined at this stage (Fig. 6H). The distributions of lacZ-positive cells in the Dom/b3IIIa-lacZ hybrid mutant embryos from 12.5 to 16.5 dpc are summarized in Figure 7B. Compared with the lacZ expression patterns of b3IIIa-lacZ embryos in the wild-type background, it is clear that, in the Dom mutant background, the lacZ staining was reduced and the migration of lacZ-positive cells was retarded (Fig. 7A). Our analysis here further supports that the b3IIIa enhancer activity was specific for the vagal neural crest cells, which migrated into the gastrointestinal tract, proliferated, and differentiated to form the enteric nervous system.
b3IIIa-lacZ Transgene Marks the Developing ENS
We have shown that the b3IIIa-lacZ transgene is specifically expressed at 9.5 dpc in the vagal neural crest that migrates down to the foregut region. At 10.5 dpc, the lacZ-marked cells appear as a migrating wave from the foregut to reach the distal hindgut by 14.5 dpc. After this stage, the transgene expression pattern in the gut is maintained at 16.5 dpc and until after birth (Fig. 4 and data not shown). The colonization of the developing gut by these lacZ-expressing cells was highly consistent with previous observations of the vagal neural crest using other cellular and molecular markers (Gershon et al., 1993; Gershon, 1997; Young et al., 2000; Young and Newgreen, 2001).
The specificity of the b3IIIa enhancer in directing gene expression in the developing ENS is supported by genetic analysis using the Dom mutant, a spontaneous mouse model of Hirschsprung's disease characterized by the absence of enteric ganglia in the distal gut. The pathogenesis of aganglionosis in Dom has been attributed to a mutation of the Sox10 gene that caused the failure of the neural crest to completely colonize the distal intestine (Kapur et al., 1996; Herbarth et al., 1998; Southard-Smith et al., 1998, 1999; Kapur, 1999). In the Dom/b3IIIa-lacZ hybrid mutant mice that we produced, at 12.5 dpc, the lacZ-expressing cells did not progress beyond the cecum and never colonized the large intestine. At later stages (14.5 dpc and 16.5 dpc), some lacZ-marked cells could be found in the large intestine but the density of lacZ staining was significantly reduced in the hybrid mutant embryos. The colonization of lacZ-expressing cells in the large intestine was incomplete in the distal hindgut of all the hybrid mutants (Figs. 6, 7). Our results show that the enteric neuroblasts have been marked by the lacZ cells in the Dom/b3IIIa-lacZ hybrid mutants and that the mutants displayed an absence of lacZ-marked cells in the hindgut. The b3IIIa-lacZ transgene has clearly marked the vagal neural crest cells and their ENS derivatives in the gut, further confirming the specificity of this enhancer in the developing ENS.
b3IIIa-lacZ Transgene Is Expressed in a Subset of the ENS
The vagal neural crest cells are progenitor cells that migrate rostral–caudally from the foregut to the distal hindgut, proliferate en route, populate the entire gut, and give rise to enteric neurons and glia of the ganglionic plexuses in the ENS. Whereas the temporal and spatial distribution of the lacZ-expressing cells closely resembles that of vagal neural crest cells, it is not clear whether the intensity of lacZ staining and density of lacZ-positive cells along the developing gut (Figs. 4, 7A) also represent the characteristics of vagal neural crest cells. It is possible that the lacZ-staining patterns reflect the b3IIIa enhancer activities in different regions of the developing gut, the activities of the enhancer could be modulated among the cells of the ENS during development.
Several cellular and molecular markers have been identified for the migrating neural crest cells as they colonize the gut, including markers for neural crest cells, neurotrophic factor receptors, neurotransmitters, and transcription factors. However, whether these markers identify all or just subsets of the ENS precursors remain controversial issues. The expression of certain markers may identify specific lineages of neural crest-derived cells that are restricted in their ultimate developmental potential. Several early markers such as Ret, Phox2b, p75NTR, and PGP9.5 are expressed in early neural crest progenitor cells that migrate into the gut, and later, these markers are expressed in the committed enteric neuroblasts (Young et al., 1998, 1999, 2003; Sidebotham et al., 2001; Young and Newgreen, 2001; Natarajan et al., 2002). To examine the lineage specificity of the b3IIIa enhancer element, we used p75NTR and PGP9.5 to mark the lacZ-expressing cells in the b3IIIa-lacZ transgenic embryos. Immunohistochemical analysis of 12.5 dpc and 14.5 dpc embryonic gut sections revealed that the lacZ-marked cells were distributed as a ring in the myenteric plexus region on the serosal side of the circular muscle but not in the submucosal region. The overall location and distribution pattern of b3IIIa-lacZ transgene-expressing cells are consistent with that of p75NTR and PGP9.5 expression, which marks the enteric neuroblasts of the myenteric plexus of the gut (McKeown et al., 2001). However, there are some cells of the myenteric plexus that are positive for p75NTR and PGP9.5 but do not express the lacZ transgene (Fig. 4E–G). Therefore, the b3IIIa-lacZ transgene is expressed only in a subset of the enteric neuroblasts in the developing gut between 12.5 dpc to 14.5 dpc. It is possible that the b3IIIa enhancer is active in a subset of enteric neuroblasts of specific sublineage.
It has been suggested that enteric neuronal precursors have already started differentiating into neurons by the time when neuronal precursors have colonized the entire gut (Payette et al., 1984). We have used a late neuronal marker, NF-M antibody (2H3), which marks differentiated neurons, to examine the gut sections of b3IIIa-lacZ transgenic embryos. At earlier stages (12.5 dpc) when the neural crest have not yet fully differentiated into enteric neurons, few of the lacZ-expressing cells, presumably those newly differentiated immature neurons, were NF-M–positive. At later stages (14.5 dpc to 16.5 dpc) when most of neural crest cells had differentiated into enteric neurons, lacZ staining overlapped extensively with the NF-M immunoreactivity but were hardly detected in the GFAP-positive population of differentiated glial cells. The b3IIIa enhancer element appeared to have specific activities in the neuronal lineage but not the glial lineage of the developing enteric nervous system. In summary, we have shown that the b3IIIa-lacZ transgene is activated in a subset of enteric neuroblasts in the gut during early development and that the transgene expression is maintained in differentiated neurons at later stages. To facilitate further characterization of the activity the b3IIIa enhancer in specific sublineages of the developing ENS, using this b3IIIa enhancer to link up with the Cre recombinase gene and generate a b3IIIa-Cre mouse line will allow the use of Cre recombinase expression as a cell lineage tracer in reporter mice such as ROSA26 (Soriano, 1999) and Z/EG (Novak et al., 2000).
b3IIIa Enhancer Is Specific for Vagal Neural Crest but Not Sacral Neural Crest
The use of tissue-specific promoters in transgenic mouse analysis to study the development of the ENS has been described previously (Kapur et al., 1992, 1996; Young and Newgreen, 2001). Among the first recognized markers of ENS precursors were the enzymes specifically involved in catecholamine biosynthesis, tyrosine hydrolase, and dopamine-β-hydroxylase (DβH). By establishing a DβH-lacZ transgenic mouse line that carried the lacZ reporter gene directed by the DβH gene promoter, it was possible to use the reporter gene to mark the enteric neuroblasts during the course of ENS development. However, the DβH promoter is not restricted to the enteric nervous system and is not vagal neural crest-specific. Using a similar lacZ reporter approach, Wnt1-lacZ was also used to mark neural crest cells from the whole range of the neural tube (Echelard et al., 1994). Recently, an ENS-specific enhancer has been identified in the Ednrb gene, lacZ expression could be detected specifically in the ENS precursors of transgenic reporter mice (Zhu et al., 2004). However, similar to other neural crest markers, the expressions of Wnt1-lacZ and Ednrb-lacZ are not limited to the vagal neural crest but include the neural crest of the body trunk.
Unlike most other neural crest-specific markers, the b3IIIa-lacZ transgene marks only the vagal region, other posterior trunk neural crest and sacral neural crest do not express the transgene. Whereas the expression of the b3IIIa-lacZ transgene in the neural tube is transient (Fig. 3), transgene expression in the ENS of the gut is persistent throughout embryonic development. There have been a lot of studies on the relative contributions of vagal and sacral neural crest to the formation of the ENS in mammalian and avian species (Burns and Le Douarin, 1998, 2001; Burns et al., 2000; Erickson and Goins, 2000; Hearn and Newgreen, 2000; Kapur, 2000). In megacolon mouse models and in Hirschsprung's disease of human, it has been known that the intestinal aganglionosis phenotype is due to abnormal vagal neural crest development, and the defective ganglia cannot be compensated by sacral neural crest cells (Burns et al., 2000). The genes involved in Hirschsprung's disease, such as RET and SOX10, are expressed in the neural crest cells of the entire body axis but not restricted to the vagal region. As yet, the tissue-specific regulatory regions of these genes have not been identified. Therefore, it has been difficult to address the problem regarding the relative contribution of vagal and sacral neural crest cells to the development of the ENS by molecular or genetics approaches. In this study, we demonstrated that the regional specificity provided by this Hoxb3 enhancer element makes the b3IIIa enhancer a unique molecular tool for generating specific mouse models in addressing the differential contribution of vagal and sacral neural crest cells to the ENS.
Regulation of b3IIIa Enhancer Activity
Hox genes are widely expressed in the developing gut, but the roles of Hox genes in the development of the enteric nervous system are not clearly understood. We have shown that the Hoxb3 gene is expressed in the developing enteric nervous system (Fig. 1). The endogenous expression of Hoxb3 in the ENS is likely to be regulated by the b3IIIa enhancer. Our previous transgenic analysis of the Hoxb3 IIIa enhancer element has shown that this element controls gene expression in both the neural tube and the associated neural crest cells. We have demonstrated that the neural expression domains directed by this enhancer element were regulated by Hox genes, but the trans-acting factors for controlling the vagal neural crest expression have not yet been identified (Yau et al., 2002). To date, several transcription factors such as Sox10, Mash1, Phox2a and Phox2b, dHand, Hox11L1, and Hoxb5 have been found to be important for the control of mammalian ENS development (Gershon, 1997; Gariepy, 2001; Young and Newgreen, 2001; Newgreen and Young, 2002). Some of these factors could be candidate upstream regulators of the b3IIIa enhancer, or they could be downstream effectors involved in the Hoxb3 regulatory pathway. However, except for Sox10, the binding sites for most of these transcription factors have not been clearly defined; the molecular interacting pathways for controlling the b3IIIa activity have yet to be determined.
As abnormal development of the enteric neural crest is the underlying cause for the intestinal aganglionosis condition in Hirschsprung's disease in humans, studying the control of vagal neural crest development will have significant implications in understanding the disease mechanism. Currently, mutations of several genes (RET, GDNF, GFR1, EDNRB, EDN3, ECE-1, NTN, SOX10, SIP-1) (Amiel and Lyonnet, 2001; Passarge, 2002; Iwashita et al., 2003) are implicated as the cause of Hirschprung's disease. However, the molecular mechanisms controlling the expression of these genes have not been determined. We have established in this study the b3IIIa-lacZ transgene as a specific marker for vagal neural crest and the enteric nervous system in the development gut. The vagal neural crest-specific activity of the b3IIIa enhancer element makes it a unique molecular tool for disease gene manipulation and in delivery of therapeutic molecules in transgenic animal model studies.
In Situ Hybridization Analysis
Expression of Hoxb3 in the developing gut was examined by in situ hybridization using digoxigenin (DIG) -labeled riboprobes on mouse embryo sections using procedures similar to that described previously (Wilkinson et al., 1992). DIG-labeled riboprobes were synthesized using DIG RNA labeling kit (Roche) according to manufacturer's protocol. The Hoxb3 probe used was a 700-bp BamHI–HindIII fragment containing part of the coding and 3′-untranslated regions (Fig. 2A). For the Ret gene, the probe was derived from exon 2 to exon 6 of the mouse Ret 9 cDNA (Lee et al., 2003). After the hybridization process and final posthybridization washing, hybridization signals were amplified with the Tyramide Signal Amplification System and Streptavidin–horseradish peroxidase conjugate following manufacturer's protocol (TSA Biotin System, PerkinElmer). Sections were counterstained lightly in hematoxylin, dehydrated in alcohol, cleared in xylene, and mounted in DPX mountant (BDH).
Transgenic and Dom Mutant Mice
The transgenic b3IIIa-lacZ reporter construct used was as described in Yau et al. (2002). The enhancer element IIIa is a SacI–StuI fragment of the Hoxb3 element III (Kwan et al., 2001; Yau et al., 2002). The vagal neural crest-specific activity is restricted to a 482-bp Sau3A fragment in the IIIa enhancer (Fig. 2A). Construct b3IIIa-lacZ was generated by cloning the 482-bp Sau3A fragment into a lacZ reporter cassette pB4ZA (Fig. 2A). The DNA insert was released from the plasmid vector by XhoI digestion, and the DNA was purified and used for microinjection using fertilized oocytes obtained from F1 (CBA X C57BL/10) mice. Transgenic mice were genotyped by polymerase chain reaction (PCR) using DNA extracted from yolk sac or tail biopsy. Polymerase chain reactions were performed using primers 26P (5′-GAAATTAATGGCTATGAGTTCCTT-3′) and BG3Z (5′-ATGGGCGCATCGTAACCGTGCAT-3′).
Dom was obtained from the Jackson Laboratory (Bar Harbor, ME) and was maintained in a hybrid background of C57BL/6J x C3HeB/FeJ. To genotype the Dom embryos, the microsatellite marker D15Mit71 was used as a linkage marker for the Dom allele. PCR was performed using the primers D15MIT71-F (5′-CCCAACTCATATGTATTATCCTGC-3′) and D15MIT71-R (5′-TAATGACAGTGCCAAATCTTGG-3′), which generated a 14-bp polymorphism between C57BL/6J and C3HeB/FeJ in the microsatellite sequence that marked the Sox10 locus (Herbarth et al., 1998). The Dom mutants were crossed with the b3IIIa-lacZ transgenic mice to produce double heterozygous mutant embryos that carried the Dom and the b3IIIa-lacZ alleles.
Examination of lacZ Transgene Expression Patterns
The expression patterns of the lacZ reporter gene in the transgenic embryos were analyzed by whole-mount β-galactosidase activity staining as described in Kwan et al. (2001). The embryos or isolated gastrointestinal tracts were fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) at 4°C for 30 min to 2 hr, depending on the size of the specimen. The fixed embryos were then washed in PBS with 0.02% NP40 at room temperature three times for 30 min each. The embryos were stained in the dark in X-gal staining solution containing 1 mg/ml X-gal, 5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, 2 mM MgCl2, 0.01% sodium deoxycholate, and 0.02% NP40 in PBS. The stained embryos or gastrointestinal tracts were rinsed with PBS and post-fixed in 4% paraformaldehyde at 4°C. Images of the whole-mount X-gal–stained embryos or gastrointestinal tracts were taken on a stereomicroscope fitted with a Sony digital camera under darkfield or brightfield illumination.
Immunohistochemical Analysis of the Embryonic Gut
Paraffin-embedded sections of transgenic embryos previously stained with X-gal were observed directly for analysis of lacZ expression patterns and also examined using antibodies against neurofilament NF-M (2H3; Developmental Studies Hybridoma Bank), PGP9.5 (Biogenesis), p75NTR (Chemicon), and GFAP (Dako) similar to that described in Lui et al. (2001). For preparation of paraffin sections, X-gal–stained, post-fixed embryos were dehydrated and embedded in paraffin wax. Paraffin sections (6 μm) were collected on glass slides coated with TESPA (Sigma, St. Louis, MO). The slides were preincubated for 30 min in 0.3% hydrogen peroxide before antibodies were applied onto the sections overnight. The sections were then washed several times with PBS before incubation with biotinylated secondary antibodies (Dako) for 1 hr. After several rinses with PBS, the ABC immunodetection kit (Vector Laboratories) was used to visualize the signals using diaminobenzidine substrate (Sigma) according to the manufacturer's instruction. After color developing, the sections were rinsed with water, dehydrated, counterstained with hematoxylin and mounted with Permount (Fisher Scientific).
We thank Sheila S.L. Tsang for animal husbandry. This project is supported by a Hong Kong RGC grant to M.H.S.
- 2001. Hirschsprung disease, associated syndromes, and genetics: a review. J Med Genet 38: 729–739. , .
- 1999. Loss of Hoxa5 gene function in mice perturbs intestinal maturation. Am J Physiol 277: C965–C973. , , , .
- 2002. Homeobox genes in gut development. Gut 51: 450–454. .
- 1998. The sacral neural crest contributes neurons and glia to the post-umbilical gut: spatiotemporal analysis of the development of the enteric nervous system. Development 125: 4335–4347. , .
- 2001. Enteric nervous system development: analysis of the selective developmental potentialities of vagal and sacral neural crest cells using quail-chick chimeras. Anat Rec 262: 16–28. , .
- 2000. Sacral neural crest cells colonise aganglionic hindgut in vivo but fail to compensate for lack of enteric ganglia. Dev Biol 219: 30–43. , , .
- 1994. Fetal endoderm primarily holds the temporal and positional information required for mammalian intestinal development. J Cell Biol 126: 211–221. , , , .
- 1997. Changing intestinal connective tissue interactions alters homeobox gene expression in epithelial cells. J Cell Sci 110(Pt 11): 1317–1324. , , , , , .
- 1996. Common origin and developmental dependence on c-ret of subsets of enteric and sympathetic neuroblasts. Development 122: 349–358. , , , , .
- 1994. Cis-acting regulatory sequences governing Wnt-1 expression in the developing mouse CNS. Development 120: 2213–2224. , , .
- 2000. Sacral neural crest cell migration to the gut is dependent upon the migratory environment and not cell-autonomous migratory properties. Dev Biol 219: 79–97. , .
- 2003. HOXB5 expression is spatially and temporarily regulated in human embryonic gut during neural crest cell colonization and differentiation of enteric neuroblasts. Dev Dyn 228: 1–10. , , , , .
- 2001. Intestinal motility disorders and development of the enteric nervous system. Pediatr Res 49: 605–613. .
- 1997. Genes and lineages in the formation of the enteric nervous system. Curr Opin Neurobiol 7: 101–109. .
- 1993. From neural crest to bowel: development of the enteric nervous system. J Neurobiol 24: 199–214. , , .
- 2000. Endoderm development: from patterning to organogenesis. Trends Genet 16: 124–130. , .
- 2000. Lumbo-sacral neural crest contributes to the avian enteric nervous system independently of vagal neural crest. Dev Dyn 218: 525–530. , .
- 1998. Mutation of the Sry-related Sox10 gene in Dominant megacolon, a mouse model for human Hirschsprung disease. Proc Natl Acad Sci U S A 95: 5161–5165. , , , , , , , , .
- 1998. A conserved retinoic acid responsive element in the murine Hoxb-1 gene is required for expression in the developing gut. Development 125: 3235–3246. , , , .
- 2003. Hirschsprung disease is linked to defects in neural crest stem cell function. Science 301: 972–976. , , , , .
- 1999. Early death of neural crest cells is responsible for total enteric aganglionosis in Sox10(Dom)/Sox10(Dom) mouse embryos. Pediatr Dev Pathol 2: 559–569. .
- 2000. Colonization of the murine hindgut by sacral crest-derived neural precursors: experimental support for an evolutionarily conserved model. Dev Biol 227: 146–155. .
- 1992. A transgenic model for studying development of the enteric nervous system in normal and aganglionic mice. Development 116: 167–175. , , .
- 2004. The influence of Hox genes and three intercellular signalling pathways on enteric neuromuscular development. Neurogastroenterol Motil 16(Suppl 1): 8–13. , , , .
- 1996. Abnormal microenvironmental signals underlie intestinal aganglionosis in Dominant megacolon mutant mice. Dev Biol 174: 360–369. , , , , , .
- 2001. Regulatory analysis of the mouse Hoxb3 gene: multiple elements work in concert to direct temporal and spatial patterns of expression. Dev Biol 232: 176–190. , , , .
- 1984. Association of megacolon with a new dominant spotting gene (Dom) in the mouse. J Hered 75: 435–439. , .
- 2003. 3′ Splicing variants of ret receptor tyrosine kinase are differentially expressed in mouse embryos and in adult mice. Biochim Biophys Acta 1627: 26–38. , , , , .
- 2001. CDX-1 and CDX-2 are expressed in human colonic mucosa and are down-regulated in patients with Hirschsprung's disease associated enterocolitis. Biochim Biophys Acta 1537: 89–100. , , , .
- 2001. Development of the submucous plexus in the large intestine of the mouse. Cell Tissue Res 303: 301–305. , , .
- 2002. Requirement of signalling by receptor tyrosine kinase RET for the directed migration of enteric nervous system progenitor cells during mammalian embryogenesis. Development 129: 5151–5160. , , , .
- 2002. Enteric nervous system: development and developmental disturbances—part 1. Pediatr Dev Pathol 5: 224–247. , .
- 2000. Z/EG, a double reporter mouse line that expresses enhanced green fluorescent protein upon Cre-mediated excision. Genesis 28: 147–155. , , , , .
- 2002. Dissecting Hirschsprung disease. Nat Genet 31: 11–12. .
- 1984. Neurofilament expression in vagal neural crest-derived precursors of enteric neurons. Dev Biol 105: 273–287. , , .
- 1999. Coordinated expression of 3′ hox genes during murine embryonal gut development: an enteric Hox code. Gastroenterology 117: 1339–1351. , , , .
- 1995. Sonic hedgehog is an endodermal signal inducing Bmp-4 and Hox genes during induction and regionalization of the chick hindgut. Development 121: 3163–3174. , , , , , .
- 1998. Epithelial-mesenchymal signaling during the regionalization of the chick gut. Development 125: 2791–2801. , , , .
- 1998. Region-specific expression of murine Hox genes implies the Hox code-mediated patterning of the digestive tract. Genes Cells 3: 51–64. , , , , , , , , .
- 1992. Analysis of the murine Hox-2.7 gene: conserved alternative transcripts with differential distributions in the nervous system and the potential for shared regulatory regions. EMBO J 11: 1825–1836. , , , , , , .
- 2001. Assessment of protein gene product 9.5 as a marker of neural crest-derived precursor cells in the developing enteric nervous system. Pediatr Surg Int 17: 304–307. , , , , , .
- 1999. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet 21: 70–71. .
- 1998. Sox10 mutation disrupts neural crest development in Dom Hirschsprung mouse model. Nat Genet 18: 60–64. , , .
- 1999. The Sox10(Dom) mouse: modeling the genetic variation of Waardenburg- Shah (WS4) syndrome. Genome Res 9: 215–225. , , , , , , .
- 1993. Structural abnormalities associated with congenital megacolon in transgenic mice that overexpress the Hoxa-4 gene. Dev Dyn 198: 28–53. , , , , , , .
- 1998. Fetal development of the enteric nervous system of transgenic mice that overexpress the Hoxa-4 gene. Dev Dyn 211: 269–291. , , , , .
- 1997. 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 124: 4781–4791. , , , , .
- 1992. Whole mount in situ hybridisation of vertebrate embryos. In: WilkinsonDG, editors. In situ hybridisation: a practical approach. Oxford, UK: Oxford IRL Press. p 75–83. .
- 1989. Transgenic mice overexpressing the mouse homeobox-containing gene Hox-1.4 exhibit abnormal gut development. Nature 337: 464–467. , , , , .
- 2002. Auto/cross-regulation of Hoxb3 expression in posterior hindbrain and spinal cord. Dev Biol 252: 287–300. , , , , , .
- 1995. Coordinated expression of Abd-B subfamily genes of the HoxA cluster in the developing digestive tract of chick embryo. Dev Biol 169: 76–89. , , .
- 2001. Enteric neural crest-derived cells: origin, identification, migration, and differentiation. Anat Rec 262: 1–15. , .
- 1998. A single rostrocaudal colonization of the rodent intestine by enteric neuron precursors is revealed by the expression of Phox2b, Ret, and p75 and by explants grown under the kidney capsule or in organ culture. Dev Biol 202: 67–84. , , , , , .
- 1999. Expression of Ret-, p75(NTR)-, Phox2a-, Phox2b-, and tyrosine hydroxylase-immunoreactivity by undifferentiated neural crest-derived cells and different classes of enteric neurons in the embryonic mouse gut. Dev Dyn 216: 137–152. , , , .
- 2000. Embryology and development of the enteric nervous system. Gut 47(Suppl 4): iv12–iv14. , , .
- 2003. Acquisition of neuronal and glial markers by neural crest-derived cells in the mouse intestine. J Comp Neurol 456: 1–11. , , .
- 2004. Dynamics of neural crest-derived cell migration in the embryonic mouse gut. Dev Biol 270: 455–473. , , , , , , .
- 1999. Hox genes and the making of sphincters. Nature 401: 761–762. , .
- 2004. Spatiotemporal regulation of endothelin receptor-B by SOX10 in neural crest-derived enteric neuron precursors. Nat Genet 36: 732–737. , , , , , .