The repulsive guidance molecule (RGM), a GPI-anchored and membrane-associated glycoprotein, has been cloned and at first functionally characterized as a molecular determinant for the retinotectal map formation of the chick embryo (Monnier et al., 2002). Recombinant RGM at low nanomolar concentration induced a collapse of temporal but not of nasal growth cones and guided temporal retinal axons in vitro by a repulsive guiding activity. The molecular structure of RGM includes a N-terminal signal sequence, a RGD motif, a partial von Willebrand factor type D domain, and a C-terminal GPI-anchor domain. Three genes have been isolated in the mouse genome and the overall sequence homologies of chick RGM to mouse RGMa, RGMb, and RGMc are 78, 43, and 40%, respectively (Schmidtmer and Engelkamp, 2004). All three murine members of this protein family share no significant sequence homology with any other known guidance molecules. The expression patterns of the three homologues were identified in the mouse embryo by in situ hybridization (Oldekamp et al., 2004; Schmidtmer and Engelkamp, 2004). Thus, RGMa and RGMb were predominantly expressed in the developing and adult central nervous system (CNS) in complementary patterns with little overlap. Subtype-specific mRNA was detected in the telencephalon, diencephalon, midbrain, hindbrain, and spinal cord. RGMb could also be seen in the peripheral nervous system, e.g., in cranial nerve ganglia, dorsal root ganglia, and cochlear ganglion. Interestingly, both subtypes appear in fetal entodermal tissues including the lung and cells in the outer layers of the digestive tract. RGMb has additional expression domains in the pancreas. RGMc shows expression restricted to all striated muscles and the myocardium.
Meanwhile, neogenin, a former netrin-1 receptor, was demonstrated to function also as a RGM receptor (Rajagopalan et al., 2004). Neogenin was first identified in chicken as a highly regulated protein in the developing nervous system and gastrointestinal tract (Vielmetter et al., 1994). The human homolog is roughly 50% identical to the protein deleted in colorectal cancer (DCC), a candidate tumor suppressor that is involved in gut tumor progression and also in neural development (Keino-Masu et al., 1996; Gad et al. 1997; Mazelin et al., 2004). A study in chick embryos identified RGM and its receptor neogenin in regulating neuronal cell survival (Matsunaga et al., 2004). Thus, neogenin is a dependence receptor inducing cell death in the absence of RGM, whereas the presence of RGM inhibits this effect. In the mouse, RGMa is involved in adhesion of dorsal root ganglia cells (Samad et al., 2004), in neural tube closure (Niederkofler et al., 2004), and in the formation of afferent connections in the dentate gyrus (Brinks et al., 2004). Recently, we discovered that in addition to RGMa, RGMb also binds to neogenin and regulates proliferation, migration, and differentiation of dentate gyrus precursor cells (Conrad et al., unpublished data).
In contrast to RGMs, several studies have demonstrated that other guidance cues and their receptors are involved in the development of the intestinal epithelium (Batlle et al., 2002; Mazelin et al., 2004) and the peripheral autonomic nervous system (reviewed by Young et al., 2004). The guidance molecule netrin-1 and its receptor DCC mediate the centripetal migration of ganglia cells from the myenteric to the submucosal region of the gut (Jiang et al., 2003). Netrin-1 and DCC are also implicated in colorectal tumorigenesis and apoptosis (Mazelin et al., 2004). Ephrins are another class of guidance molecules that have been reported to be involved in colorectal cancer progression and migration of intestinal paneth cells (Batlle et al., 2002).
The expression pattern and specific biological function of RGM in the adult intestine is still unknown. Therefore, we investigated in detail the expression of the repulsive guidance molecules (RGMs) during fetal and postnatal development of the mouse gut using RT-PCR, in situ hybridization, Western blot, and immunocytochemistry.
In this study, we investigated the expression of RGM subtypes during development in the mouse gut by RT-PCR, in situ hybridization, Western blot, and immunocytochemistry.
By using RT-PCR experiments, we analyzed mRNA expression of RGMs and neogenin in embryonic and postnatal gut samples (E14, E18, P0, P7, P14, P21, and adult). RGMb and neogenin were expressed in all investigated developmental stages. RGMa was first clearly detected with embryonic stage E18 whereas RGMc subtype was absent at all (Fig. 1). A more sensitive real-time quantitative RT-PCR analysis demonstrated that subtype RGMa mRNA was already very weakly detectable at E14 and expression significantly increased 3.5-fold at E18 (data not shown).
In order to analyze cell type specific expression of RGMs during gut development, we carried out in situ hybridization for all RGM subtypes. Apart from the published RGM expression pattern during development of the CNS (Oldekamp et al., 2004; Schmidtmer and Engelkamp, 2004), subtypes RGMa and RGMb but not RGMc were also detectable in the developing and adult gastrointestinal tract. In the mouse embryo (E14), both subtypes RGMa (Fig. 2A,C) and RGMb (Fig. 2B,D) were expressed in the wall of the digestive tract.
In the neonatal gut, RGMa and RGMb expression was located in enteric ganglia cells. This is shown exemplarily for subtype RGMb in the small intestine (Fig. 3A) and in the colon (Fig. 3B). Similar to the neonatal stage, RGMa and RGMb were expressed in the ganglia cells of the adult small intestine (Fig. 3C,E) and colon (Fig. 3D,F) between the inner circular and the outer longitudinal muscle layers (plexus myentericus) and in ganglia cells of the submucosa (plexus submucosus). The combination of RGMa in situ hybridization with immunohistochemistry for neurofilament demonstrated that RGM is expressed in enteric neurons (Fig. 4A,B).
In contrast to ganglia cells, RGMa and RGMb expression in intestinal epithelial cells firstly occurred at postnatal day 7 (data not shown) and was maintained until the adult stage (RGMb: Fig. 3C,D; RGMa: Fig. 3E,F). In negative controls using a RGMa sense probe, no signals were detected (Fig. 3G,H).
Interestingly, expression of RGMa and RGMb was predominantly observed in the lower part of the intestinal crypt compartment. In the same compartment, proliferating epithelial precursors and stem cells are localized. However, in the crypt base of the small intestine RGMa was also expressed in non-proliferating Paneth cells. This was shown in co-staining experiments by RGMa in situ hybridization and immunohistochemistry for lysozyme, an enzyme strongly expressed in this cell type (Fig. 4D,E).
In addition to mRNA analysis, we investigated RGMa protein expression in the gut. Tissue samples were stained by indirect immunofluorescence using the polyclonal rabbit RGMa antibody (Monnier et al. 2002). The staining in colonic tissue was generally weaker as observed in the small intestine (data not shown).
Similar to in situ hybridization results, RGMa protein was predominantly expressed in crypt cells of the intestinal epithelium (Fig. 5A) and in ganglia cells (Fig 5B). In control experiments lacking the primary antibody, no staining was observed in intestinal crypt cells and in the enteric nervous system whereas some apical cells of the villi, smooth muscle cells of the outer longitudinal muscle layer and cells of the serosa showed an unspecific staining (data not shown).
In addition to immunocytochemistry, we analyzed the RGMa protein expression by SDS-page and Western blotting of cell homogenates prepared from embryonic and postnatal gut samples at the stages E18, P0, P7, P14, P21, and adult. At all stages, one strong protein band with a slight increase in intensity during gut development was detected at about 42 kDa, which is in accordance with data published for the mouse RGMa protein (Niederkofler et al., 2004) (Fig. 5C).
During development, a number of attractive and repulsive guidance cues are involved to achieve the complex pattern of the vertebrate central nervous system (reviewed in Dickson, 2002; de Castro, 2003). Apart from their role in the central nervous system, guidance cues are also involved in the development of the peripheral autonomic nervous system (reviewed in Young et al., 2004) and the enteric nervous system (Seaman et al., 2001; Natarajan et al., 2002, De Bellard et al., 2003, Jiang et al., 2003).
In this study, we could identify members of the RGM gene family in enteric ganglia cells during the development of the whole gut. Subtypes RGMa and RGMb were expressed in ganglia cells of the plexus myentericus and plexus submucosus whereas RGMc expression was absent. The RGM receptor neogenin is closely related to the netrin-1 receptor and tumor suppressor gene product DCC (Vielmetter et al., 1996; Keeling et al., 1997). In addition to the sequence homologies of both receptors, there seem to be developmental and functional parallels. First, these two molecules share similar tissue distribution, being expressed both in neuronal and nonneuronal (e.g., gastrointestinal) tissues (Fearon et al., 1990; Hedrick et al., 1992; Gad et al., 1997). Second, the expression of both of these receptor molecules is compatible with their involvement in terminal differentiation of specific cell types (Vielmetter et al., 1994; Mazelin et al., 2004; Niederkofler et al., 2004). Thus, it remains plausible that functional interactions may exist in a complementary way between the ligands netrin and RGM.
With this study, we could additionally show that subtypes RGMa and RGMb were expressed in the proliferating compartment of the postnatal, but not fetal and neonatal, intestinal epithelium. The intestinal epithelium is one of the most actively renewing tissues with a high regeneration capacity. Crypts of the small intestine develop during the early postnatal period (days 1–5) from the flat inter-villus epithelium, and mitotic cells segregate in these structures (reviewed in Clatworthy and Subramanian, 2001). The process of complete cyto-differentiation is completed 20 days after weaning. Continued tissue function is accomplished by a delicate balance of cell removal through apoptosis and replacement via proliferation.
Recent studies have demonstrated that guidance cues can influence epithelial morphogenesis (Batlle et al., 2002; Srinivasan et al., 2003, Yebra et al., 2003; Liu et al., 2004). The guidance molecule netrin-1 and its receptor DCC were recently implicated in colorectal tumorigenesis by regulating apoptosis (Mazelin et al., 2004). The ephrins are another class of guidance molecules that have been reported to possess an in vitro activity similar to RGMa and RGMb, causing growth cone collapse and preferential guidance of neuronal axons (Himanen and Nikolov, 2003), and have also been described to be involved in colorectal cancer progression. Batlle et al. (2002) have shown that in EphB2/EphB3 null mice, the proliferative and differentiated populations intermingle in the small intestine. Further, in adult EphB3 knock out mice, Paneth cells do not follow their downward migratory path, but scatter along crypt and villus (Batlle et al., 2002).
The expression pattern of repulsive guidance molecules in the gut gives the first evidence that RGMa and RGMb proteins may also be involved in the differentiation processes of enteric ganglia cells and adult intestinal epithelium.
Mouse embryos, embryonic and postnatal small and large intestines, were obtained from C57BL/6 mice. Animals were killed by carbon dioxide incubation. Tissue samples at embryonic stages E14, E18, postnatal stages P0, P7, P14, P21, and adult were prepared, cleaned free of mesentery, and washed with Hanks' balanced salt solution. To prevent RNA degradation, tissues were immediately frozen in liquid nitrogen and stored at −80°C until use.
In Situ Hybridization
For in situ hybridization analysis, pieces of small and large intestines were directly frozen in O.C.T. compound (Tissue-Tek; Sakura Finetek, Torrance, CA). Sections of 12 μm were sliced on a freezing microtome, mounted on silane-coated slides (2% 3-aminopropyltriethoxy-silane (Sigma Taufkirchen, Germany) in acetone), immediately dried at 60°C for 20 s, and fixed in 4% phosphate buffered paraformaldehyd for 20 min. Hybridization was performed with digoxigenin-labeled RNA probes directed against mRGMa, mRGMb, and mRGMc. A sense probe served as control. Probe generation and in situ hybridization procedure were performed as described previously (Oldekamp et al., 2004). Briefly, the procedure includes proteinase K digestion (2 μg/ml proteinase K (Roche, Mannheim, Germany) in 100 mM Tris-HCl, 50 mM EDTA, pH 8.0, for 15 min at room temperature and acetylation with TEA buffer (0.1 M triethanolamine, pH 8.0) containing 0.25% (v/v) acetic anhydride (Sigma, St. Louis, MO) twice for 5 min. After pre-hybridization with hybridization buffer (50% formamide [Sigma], 10% dextran sulfate, 5 mM EDTA, 20 mM Tris, pH 8.0, 10 mM DTT, 1x Denhardt's solution, 0.05% tRNA, 300 mM NaCl) for 1 h at 65°C, slides were incubated with fresh hybridization buffer containing the denatured DIG-labelled RNA sense or antisense probe (200 ng/ml) overnight at 65°C in a moist chamber. After hybridization, the slides were extensively washed in 1× and 0.1× SSC buffer three times for 15 min at 65°C and blocked for 30 min with blocking solution (Perkin Elmer, Boston, MA) including 4% sheep serum. Detection of DIG-labelled RNA probe includes incubation with a sheep anti-DIG peroxidase-conjugated antibody solution ((Roche) 1:1,200 in blocking solution) and a subsequent tyramine- biotin based amplification step (Perkin Elmer). The slides were then incubated with neutravidine-alkaline phosphatase- conjugated antibody solution for 30 min (Pierce, Rockford, IL; 1:750 in blocking solution). The colour development was carried out with a freshly prepared substrate solution [nitroblue tetrazolium salt (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP; Roche) in Tris buffer (0.1 M Tris, pH 9.5, 0.1 M NaCl, 50 mM MgCl2) or, alternatively, Fast Red salt solution (Sigma)]. After 3 washes with PBS, the slides were rinsed in distilled water, dried and coverslipped with Kaiser's solution (Merck).
Immunohistochemistry was performed on cryosections (12 μm). The sections were fixed in 4% phosphate buffered paraformaldehyd for 20 min at room temperature, incubated with 3% Tris buffered hydrogenperoxide for 10 min to block endogenous peroxidases, and rinsed three times in Tris buffer (25 mM Tris-HCl, pH 7.5). Fixation step was omitted if immunohistochemistry was performed after in situ hybridization. After blocking with blocking solution (Tris buffer containing 0.1% BSA, 10% swine serum, and 0.3% Triton X-100) for 30 min, sections were incubated with primary antibodies diluted in Tris-buffer containing 0.1% BSA, 4% swine serum, and 0.1% Triton X-100 overnight at 4°C. The following primary antibodies were used: polyclonal rabbit antibody against the RGMa peptide EEVVNAVEDR spanning residues 279–289 (1:100; Monnier et al., 2002), rabbit anti-human lysozyme (1:400, Dako, Hamburg, Germany), and mouse anti-human neurofilament (1:100, Dako). After three washes with Tris-buffer, the sections were incubated with biotinylated secondary antibodies for 30 min at room temperature (1:400, swine anti-rabbit IgG and goat anti-mouse IgG; Dako). Afterwards, slides were again washed three times. Detection with the ABC system was carried out according to the manufacturer's recommendations (Dako). After visualisation with 3-3′diaminobenzidine (DAB, Sigma) and hydrogen peroxide (Sigma) solution, substrate reaction was stopped by washing in distilled water. The stained sections were dried, coverslipped with Kaiser's gelatine (Merck, Darmstadt, Germany), and photographed on an inverted Zeiss microscope (Axiovert, Zeiss, Jena, Germany).
Total protein extracts were isolated with the TRIZOL reagent according to the manufacturer's instructions (Invitrogen, Karlsruhe, Germany). Protein concentration was estimated using a Bio-Rad protein assay kit (Bio-Rad Laboratories, München, Germany). After denaturation for 1 min at 95°C and separation by SDS-page according to the method of Towbin et al. (1979) with 12.5% N,N'-bis-methylene-acrylamide in a Mini-Protean II chamber (Bio-Rad) for 60 min at 100 V, proteins were electroblotted to a nitrocellulose membrane (0.45-μm pore size; Schleicher&Schuell, Dassel, Germany) at 10 V overnight at 4°C. The incubation of the rabbit RGM antibody (1:1,000) was carried out at 4°C overnight. Biotinylated secondary antibody (1:500, goat anti-rabbit; Dako) was applied for 2 h, followed by a 2-h incubation with ABC solution (1:5,000, Dako) and was then visualized by the substrate-chromogen reaction with DAB/H2O2 solution (Sigma).
Total RNA was extracted by disrupting tissues in TRIZOL reagent (Invitrogen) with an Ultra-Turrax homogenizer (IKA, Staufen, Germany). To remove contaminating genomic DNA, the total sample RNA was incubated for 30 min at 37°C in a RNase-free Dnase I solution according to the supplier's protocol (Amersham Biosciences, Grassbrunn, Germany). The reaction was stopped by adding 4 mM EDTA and the samples were heated at 65°C for 10 min.
The purified samples of RNA underwent reverse transcription (RT) according to the manufacturer's protocol containing Oligo d(pT)18 mRNA primers (New England Biolabs, Frankfurt, Germany) and Superscript II reverse transcriptase enzyme (Invitrogen, La Jolla, CA). The synthesis of cDNA was carried out for 50 min at 42°C and for 10 min at 70°C to inactivate the RT enzyme. The amplification of the mRGM and the neogenin genes in the subsequent PCR was achieved by using subtype specific primer pairs.
Primer pairs (sense and antisense) were synthesized by Sigma as follows: for RGMa (455 bp): 5′-TCAGCTGCCCCCAACTACACT-3′ and 5′-TCCTCCACGGCGTTGACTACC-3′; for RGMb (460 bp): 5′-CAGCCACGGGGGAGTCAGAG-3′ and 5′-CATCCGGATAGCGAGGGTTAG-3′; for RGMc (506 bp): 5′-GGGCAGCTCTCCTTCTCCATC-3′ and 5′-CTCCTGTCCCCGCTGTTTCCTT-3′; for neogenin (536 bp): 5′-TTCTCCAGCCCGCAGTCATCT-3′ and 5′-CTTCCAGGTGGGCCATCTCTT-3′; for GAPDH (452 bp): 5′-ACCACAGTCCATGCCATCAC-3′ and 5′-TCCACCACCCTGTTGCTGTA-3′.
The PCR reaction mixture was incubated at 95°C for 5 min and 35 cycles were performed at 95°C for 30 s, (TAn) for 40 s, 72°C for 40 s, and a final cycle with a prolonged elongation time of 10 min at 72°C. The primer specific annealing temperature (TAn) was as follows: TAn (RGMa) = 58°C, TAn (RGMb) = 66°C, TAn (RGMc) = 60°C, TAn (neogenin) = 60°C, TAn (GAPDH) = 60°C. The PCR products were analyzed by standard electrophoresis on 1% agarose gels at 100 V, stained with ethidiumbromide and photographed under UV illumination. The size of each PCR product was estimated by using a 100-bp DNA ladder standard (Invitrogen). The PCR amplification of intronless sequences necessitated inclusion of additional controls. As a control for the adequacy of the PCR primers, cDNA from an E14 embryo was used as template (positive control). To control possible contaminations of Dnase-treated samples with residual genomic DNA, the RT step was omitted (negative control). A quantitative RT-PCR evaluation was performed by using the Light-Cycler RT-PCR System according to the supplier's protocol (Roche).
We thank Claudia Barthle for her helpful comments. This work was supported by a grant of the BMBF (D.20.01365), Germany.