Members of the Netrin family and their cognate receptors have been shown to play a pivotal role in the guidance of a variety of axon projections during embryonic development in both vertebrates and invertebrates. Netrin-1 is secreted from intermediate or final targets and acts as a long-range guidance cue by forming a concentration gradient along the pathway of the exploring growth cone (Hedgecock et al.,1990; Serafini et al.,1994). Netrin-1 can also act as a short-range guidance cue to affect changes in the direction of growth cone migration at specific choice points (Deiner et al.,1997). The interaction of the Netrin receptor, DCC, with Netrin-1 results in a chemoattractive response (Hedgecock et al.,1990; Keino-Masu et al.,1996) while the co-receptor complex of DCC and UNC5 responds to Netrin-1 in a chemorepulsive manner (Hedgecock et al.,1990; Leung-Hagesteijn et al.,1992; Leonardo et al.,1997; Przyborski et al.,1998).
Neogenin, a member of the DCC receptor family, was initially characterized as a Netrin receptor due to its relatively high affinity for Netrin-1 (Keino-Masu et al.,1996). To date, a Netrin-dependent function for Neogenin in axon guidance has not been demonstrated. However, members of the Repulsive Guidance Molecule (RGM) family have been identified as axon guidance ligands for Neogenin. Rajagopalan et al. (2004) have shown that Neogenin mediates the RGM-dependent chemorepulsive activity exhibited by retinal ganglion cell axons within the visual system. Recently, we have demonstrated a role for Neogenin in neural tube formation in the zebrafish, implicating Neogenin in the migration of neurectodermal cells (Mawdsley et al.,2004). Evidence is now accumulating that Neogenin is also important for the development of a variety of organ systems outside the nervous system. Neogenin has been shown to promote adhesion between the multipotent progenitor (cap) cells and the adjacent epithelial layer of the developing mammary gland in a Netrin-1-dependent manner, thereby stabilizing the progenitor cell compartment in this tissue (Srinivasan et al.,2003). In addition, Netrin-1-Neogenin interactions promote adhesion and migration of vascular smooth muscle cells during angiogenesis (Park et al.,2004). Furthermore, Kang and colleagues (2004) have shown that Netrin ligation of Neogenin induces myotube formation in vitro, suggesting a role for Neogenin in muscle differentiation. In this context, Neogenin requires the co-receptor CDO.
We have previously described the dynamic pattern of Neogenin mRNA expression within the developing central nervous system and in many non-neural embryonic tissues (Gad et al.,1997). However, the pattern of Neogenin protein localization in the embryo has not been documented. It is now becoming clear that for many guidance receptor families, receptor activity is tightly regulated at the post-translational level. Several studies have shown that high levels of DCC protein are present on migrating axons only when they are actively navigating along the Netrin gradient (Gad et al.,2000; Shu et al.,2000). Down-regulation of DCC protein on projecting axons coincides with the arrival of the axon at choice points expressing high levels of Netrin-1. Campbell and Holt (2001) have also presented evidence that rapid changes in local DCC protein levels on the growth cone are likely to be a significant factor in controlling growth cone guidance. Moreover, the extracellular domain of DCC has been shown to be the target for metalloprotease cleavage, the consequence of which is a decrease in responsiveness of spinal cord axons to Netrin-1 (Galko and Tessier-Lavigne,2000). Therefore, the presence of guidance receptor mRNA is not necessarily a good indicator of receptor activity. Here we examine the localization of Neogenin protein in the developing mouse embryo at embryonic day 14.5 (E14.5) when organogenesis is well underway.
RESULTS AND DISCUSSION
To determine Neogenin protein localization in the E14.5 embryo, we employed an anti-peptide antiserum (CT-Neog) raised against the C-terminal peptide of mouse Neogenin. In all cases, preincubation of this antiserum with the immunizing peptide resulted in the loss of immunoreactivity thereby confirming the specificity of the antiserum. Furthermore, this antiserum could specifically detect full-length Neogenin by Western blotting after transient transfection of mouse Neogenin cDNA into HEK293T cells (Fig. 1A). No cross-reactivity was detected with the closely related Netrin receptor DCC (Cooper et al.,1995). Antiserum raised to the C-terminus of mouse DCC produced completely distinct immunoreactivity patterns from those seen with the Neogenin antiserum (Seaman et al.,2001; data not shown). The CT-Neog antiserum is directed to the C-terminal amino acid sequence of Neogenin and is, therefore, able to detect all alternatively spliced forms of mouse Neogenin (Keeling et al.,1997).
Neogenin was first identified in the developing chick retina where Neogenin protein was localized to the retinal ganglion cell axons during axonal extension (Vielmetter et al.,1994). Rajagopalan et al. (2004) have subsequently shown that Neogenin acts as a repulsive guidance receptor on retinal ganglion cell axons projecting to the tectum. In accordance with these studies, we also observed Neogenin protein in the retinal ganglion cells and their axons in the nerve fibre layer at E14.5 in the mouse retina (Fig. 1B; arrow). In addition, high levels of Neogenin protein were observed in the equatorial regions of the developing lens where the anterior capsular epithelial cells are gradually incorporating into the lens proper. While Neogenin could be seen on all aspects of the lens epithelium, it was most highly concentrated on the exterior surface of the epithelium (Fig. 1B; arrowhead). Neogenin protein localization within the retina and lens parallels Neogenin mRNA expression (Gad et al.,2000).
We have also analyzed Neogenin protein localization in other embryonic sensory systems. At E14.5, the developing cochlea consists of an epithelial tube derived from the otic placode that has undergone one and a half of the two and a half turns seen in the adult. Strong Neogenin immunostaining was seen localized to the basal surface of the columnar epithelium (Fig. 1C; arrows). In contrast, Neogenin could not be detected on the apical surface of this epithelium indicating a polarized distribution of Neogenin protein during cochlea development. In addition, neurons within the cochlear ganglion adjacent to the cochlea also displayed significant levels of Neogenin (Fig. 1C).
During early olfactory neurogenesis, the primary olfactory receptor neuronal progenitors reside at the apical surface of the nasal cavity. From E11 in the mouse, these progenitors translocate to the basal surface where they begin to differentiate into secondary precursors that have the capacity to divide further before becoming committed neuroblasts (Caggiano et al.,1994; Cau et al.,2002). Neogenin protein was localized to the basal surface of the olfactory epithelium where co-staining with an antibody directed to PCNA, a marker for dividing cells, clearly demonstrated that Neogenin was present on the cycling secondary progenitors (Fig. 1E,G; arrowheads). Neogenin was not expressed by the primary progenitors at the apical surface. Cells expressing high levels of Neogenin were also observed throughout the lamina propria separating the olfactory epithelium from the border of the olfactory bulb in the rostral-most region of the telencephalon (Fig. 1E,G). Neogenin-expressing cells were present in large aggregates in the vicinity of the PNS-CNS transition zone and in the migratory mass abutting the transition zone (Fig. 1E,G; arrows). These cell aggregates are known to comprise olfactory ensheathing cells that derive from the nasal placode and migrate towards the olfactory bulb (Fairless and Barnett,2005). Close examination of the Neogenin-expressing cells within these aggregates revealed that they projected thin, Neogenin-positive cytoplasmic processes characteristic of olfactory ensheathing cells (Fairless and Barnett,2005). Therefore, Neogenin is expressed on differentiating olfactory receptor neurons and olfactory ensheathing cells known to escort olfactory receptor axons into the olfactory bulb.
Neogenin protein was also found in other sensory ganglia including the trigeminal and the dorsal root ganglia (Fig. 1H,I). In each case, Neogenin protein was localized to fibers projecting from the neurons while low levels of protein could be detected on the cell body. In addition, Neogenin was present on sensory axons projecting from the dorsal root ganglia (Fig. 1I; arrow). We have previously demonstrated that the Netrin receptor, DCC, is also present on axons projecting from the dorsal root and trigeminal ganglia and that Netrin-3 is also closely associated with these fibers (Seaman et al.,2001; Seaman and Cooper,2001). Such co-expression suggests a complex interaction between Netrin-3 and its receptors DCC and Neogenin.
Our initial study of Neogenin mRNA expression throughout embryogenesis using in situ hybridization demonstrated that Neogenin was expressed throughout the early embryo (E8.5 to E10.5). However, as embryogenesis progressed, Neogenin mRNA became restricted to specific organ systems including gut, stomach, lung, and kidney (Gad et al.,1997). Furthermore, as organogenesis proceeded Neogenin mRNA remained high in the mesenchyme while it was down-regulated in simple epithelia. Immunohistochemical analysis of Neogenin protein using the CT-Neog antiserum has allowed us to refine our knowledge of the Neogenin expression pattern during organogenesis.
At E14.5, when the three layers of the embryonic midgut have been established, Neogenin protein was present at high levels in the epithelial layer lining the lumen (Fig. 2A,C). While Neogenin immunoreactivity was seen on the apical, lateral, and basal membranes of the epithelial cells it was more intense on the basolateral surface of these cells (Fig. 2C; arrow). Vielmetter et al. (1994) have also observed Neogenin protein on the basal surface of the epithelium of the embryonic chick gut. In contrast, the surrounding mesenchyme (submucosa) did not express Neogenin. However, individual Neogenin-positive cells were seen distributed throughout the mesenchyme (Fig. 2C; arrowheads). The inner circular muscle layer adjacent to the submucosa also displayed high concentrations of Neogenin while the outer longitudinal muscle layer was Neogenin-negative (Fig. 2A). In contrast to the gut epithelium, the epithelial cells comprising the gastric mucosa did not display Neogenin on their surface (Fig. 2D). However, as in the gut, individual Neogenin-positive cells were seen throughout the stomach mesenchyme. Significant levels of Neogenin protein were also found in the smooth muscle layer of the E14.5 stomach.
In the developing pancreas, Neogenin was present on the epithelial lining of the pancreatic ducts and on the acinar cells bordering the lumen of the pancreatic buds (Fig. 2D,F). In both instances, Neogenin was again predominantly localized to the basolateral surface of the epithelial cells (Fig. 2F; arrows). Only very low levels of Neogenin were associated with the mesenchyme adjacent to the simple epithelia within the developing pancreas.
In contrast to its localization in the embryonic gut and pancreas, Neogenin protein was observed only in the mesenchymal cells of the E14.5 lung and kidney and not in the epithelium. Strong Neogenin immunoreactivity was seen on the plasma membrane of the mesenchymal cells surrounding the epithelial lining of the terminal bronchioles (Fig. 2G,H; arrowhead) and on mesenchymal cells adjacent to the epithelium of the segmental bronchi (Fig. 2G,H; arrow). Although the overall level of Neogenin expression in the E14.5 kidney was low, Neogenin-positive mesenchymal cells were seen surrounding the epithelium of the primitive glomeruli and metanephric tubules (Fig. 2J,K; arrows). Neogenin was also observed on mesenchymal cells within the adrenal gland (Fig. 2J).
Neogenin immunoreactivity was observed in differentiating muscle and cartilage throughout the E14.5 embryo. Neogenin protein was present at high to intermediate levels in the cartilaginous primordia of the head, including the primordia of the cranium, the nasal bone and septum, and the maxilla and mandible (Fig. 3A, data not shown). Neogenin was also seen at lower levels in the mesenchyme surrounding these structures. Neogenin immunoreactivity was observed within the condensing cartilage of the vertebral bodies in the spinal column (data not shown) and the cartilaginous condensations within the hindlimbs and forelimbs (Fig. 3C,D). Differentiating muscle also displayed significant levels of Neogenin in the developing limbs and face (Fig. 3C,D, data not shown). The identification of Neogenin protein in differentiating embryonic muscle supports the study of Kang et al. (2004) demonstrating that Neogenin is involved in myotube differentiation in C2C12 cells in vitro. Vascular smooth muscle cells lining blood vessels and arteries also exhibited strong Neogenin immunoreactivity throughout the embryo (Figs. 2G,H, 3F, data not shown) supporting a previous study implicating Neogenin in Netrin-1-dependent vascular smooth muscle cell migration (Park et al.,2004).
In summary, at E14.5 when organogenesis is progressing rapidly, Neogenin is restricted to distinct tissue layers within a given organ. Neogenin is present in many embryonic epithelia including those in the gut, pancreas, cochlea, and nasal cavity where it is predominantly restricted to the basal surfaces of the epithelial cells. However, little Neogenin is expressed in the mesenchyme of these tissues. Our observation that Neogenin protein is polarized to discrete subdomains of epithelial membranes suggests that Neogenin may be involved in establishing or maintaining the orientation and/or integrity of epithelial sheets by interacting with its ligand in the local environment. In support of this hypothesis, Srinivasan and colleagues have shown that Neogenin is required in the multipotent cap cells (epithelial cells) of the developing mammary gland to maintain close apposition between these cells and the Netrin-1-expressing prelumenal epithelium facing the lumen of the gland. Loss of Neogenin results in detachment and inappropriate migration of cap cells demonstrating that Neogenin is required to stabilize the interaction between cap and prelumenal cells.
In the lung and kidney, Neogenin is restricted to mesenchymal cells but is absent from the epithelia in these organs. Knockdown of Neogenin expression in early zebrafish embryos resulted in the failure of neural tube formation and somitogenesis due to cell migration defects (Mawdsley et al,.2004). Furthermore, Netrin-1 has been shown to induce vascular smooth muscle cell migration in a Neogenin-dependent manner (Park et al.,2004). Since mesenchymal cells are highly motile, it is likely that Neogenin promotes or steers mesenchymal cell migration during organogenesis. In accordance with its proposed axon guidance function, Neogenin is present on peripheral nervous system axons (trigeminal and dorsal root ganglion axons) and central nervous system axons (retinal ganglion cell axons). We have also found Neogenin protein on differentiating skeletal muscle, the smooth muscle of the stomach, blood vessels, and arteries, and condensing cartilage throughout the embryo. Therefore, Neogenin is expressed in a broad range of embryonic tissues and is likely to play a key role in establishing the morphological architecture in many developing organ systems.
The CT-Neog rabbit polyclonal antiserum was raised against a unique C-terminal peptide (VQETTRMLEDSESS) of mouse Neogenin corresponding to amino acids 1454–1467 of the published sequence (Keeling et al.,1997). For immunohistochemistry on frozen sections, CT-Neog was diluted 1:250. To demonstrate specificity, diluted antiserum was preincubated with the immunizing peptide (100 μg/ml) for 1 hour prior to addition to sections.
Tissue Preparation and Immunohistochemical Staining
E14.5 C57Bl/6 embryos were dissected from the uterine horn, and immediately flash-frozen in O.C.T. Compound (Tissue-Tek, CA). Sagittal and coronal sections (12 μm) were cut using a Leica CM3050 Cryostat and mounted onto SuperFrost Plus coated slides (Microm, Germany). The use of animals as described here was approved by the Animal Ethics Committee of the University of Queensland in accordance with the guidelines stipulated by the National Health and Medical Research Council of Australia.
Sections were post-fixed in absolute ethanol for 10 min, rehydrated in 70%, followed by 30% ethanol-PBS, and then washed in PBS. It should be noted that the C-terminal epitope recognized by this antiserum is very sensitive to fixation. The specific immunoreactivity demonstrated in this study was achieved by the use of a very mild fixative (ethanol) after cryostat sectioning without prior exposure to any other fixation protocol. Sections were pre-blocked and permeabilized for several hours at room temperature in blocking solution (PBS; 2% fetal bovine serum; 2% goat serum; 0.2% Triton X-100). Sections were incubated in primary antibodies in blocking solution overnight at 4°C. Neogenin immune complexes were detected by amplification with goat anti-rabbit IgG conjugated to biotin (1:500, Chemicon, Temecula, CA) in blocking solution at room temperature for 1 hr followed by incubation with streptavidin-conjugated to Alexa Fluor 488 (1:1,000, Molecular Probes, Eugene, OR). Mouse anti-PCNA (Sigma, St. Louis, MO) was used at 1:500 and the secondary antibody was goat anti-mouse IgG conjugated to Alexa Fluor 568 (1:1,000, Molecular Probes). In some instances 4′,6-diamidino-2-phenylindole (DAPI, 1 μg/ml), was added during this step to visualize cell nuclei. Slides were mounted in DakoCytomation Fluorescent Mounting Medium (DAKO, Carpinteria, CA). Images were acquired on an Olympus IX81 microscope using AnalySIS software, or on a Zeiss Axioplan 2 microscope using Axiovision software.
Specificity of the CT-Neog antiserum was verified by Western blotting of Neogenin-transfected HEK293T cells. Cells were transfected with cDNAs encoding full-length Neogenin (Myc-tagged at the C-terminus) or DCC (Myc-tagged at the C-terminus) using Fugene (Roche Diagnostics, IN) according to the manufacturer's instructions. Lysates were run on a Novex 4-20% Tris-Glycine gel (Invitrogen, La Jolla, CA) and transferred to Immobilon membrane (Millipore, Bedford, MA). Western blotting was performed using the CT-Neog antiserum (1:30,000), or the 9E10 anti-Myc monoclonal antibody (1 μg/ml). Finally, bound protein was detected using a horse radish peroxidase-conjugated secondary antibody (1:30,000) (BioRad, Richmond, CA) and the SuperSignal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL).
We thank Dr. Jim Pickles, Dr. Richard Anderson, and Dr. Heather Young for their scientific advice. We also thank Mr. Paul Addison and Dr. Christine Neyt for advice on the preparation of histological samples, Mr. Dennis Advani for production of the figures, and Ms. Rowan Tweedale for critical reading of the manuscript.