We isolated and cloned the chicken NOGO-A coding and 3' untranslated region. The NOGO-A sequence aligned to chromosome 3 of the chicken genome with 10 exons spanning 45 kb (Fig 1A). Sequence comparison with the chicken genome revealed the NOGO-A sequence to be lacking 248 bp within the first exon. Given that NOGO-A and -B isoforms originate from alternative splicing, we deduced our isolated sequence to be a shorter splice variant, classified as NOGO-A2. Northern analysis with a NOGO-A specific probe (see probes in Fig 1A) confirmed the secondary splice product. The larger NOGO-A1 and NOGO-B isoforms were predicted based on the chicken genome and northern hybridization. The chicken NOGO-C sequence was described previously 20.

To determine the expression profile of NOGO-A during CNS development, we used the NOGO-A specific probe for northern analysis. Hybridization with poly-A RNA isolated from chick brains of HH15 (E2.5), HH25 (E5), HH36 (E10) and HH45 (E20, just prior to hatching) revealed a 4.3 kb band (NOGO-A1), consistent with NOGO-A transcripts in other species 789, and a shorter 3.8 kb band (NOGO-A2) (Fig 1B). At the earliest stage (HH15) the longer NOGO-A1 transcript was nearly absent. Overall, combined NOGO-A expression increased with development reaching a maximum just prior to hatching (E20).

To ascertain whether NOGO isoforms were differentially expressed during development, we designed a probe to the 3' end of NOGO which was common to the NOGO-A, NOGO-B and NOGO-C isoforms and performed northern blot analysis of the same embryonic stages as above. In agreement with our previous data, 4.3 kb and 3.8 kb transcripts were found corresponding to NOGO-A1 and A2. In addition, 2.3 kb and 1.7 kb bands corresponding to the NOGO-B and NOGO-C isoforms were also identified (Fig 1C). Collectively, NOGO transcription increased with development. The NOGO-A profile demonstrated by the common probe was the same as that of the NOGO-A specific probe. The progressive changes in NOGO-B expression were similar to NOGO-A at early stages, but showed a marked increase at E20. In contrast, NOGO-C, which localizes to neurons and mesenchymal tissues, exhibited strong somewhat uniform expression throughout development. Northern analyses were quantified by densitometry of the autoradiographs and normalized to β-ACTIN levels (Fig 1B &1C-graphs).

Multiple alignment of the chicken NOGO-A shared 49%, 47%, 46% and 32% identity with human, mouse rat and xenopus (Rtn4.2-A1), respectively. The C-terminal 189aa, i.e., the common NOGO/Reticulon region, exhibited significant conservation (>75%) consistent with what has been previously reported 20. In addition, eight conserved regions (CR) were found within the NOGO-A specific coding sequence, suggesting potential functional domains (Fig 2). Conserved regions had ≥ 70% for 7 or more amino acids with a sequence identity/strong group conservation. Strong group refers to amino acids with similar chemical structure and function. Interestingly, there was no significant conservation found within the first two coding regions shared by NOGO-A and -B isoforms, although this region corresponds to a previously identified inhibitory domain in rat, NiRΔ2 11. Rather, the first conserved region (CR1) initiated at the first amino acid residue of the NOGO-A specific coding region. Additionally, CR4-6 correspond to a potent inhibitory domain, NiGΔ20 11.

The quantitative limitations of northern analysis precluded our ability to detect NOGO-A expression at the earliest stages of chick CNS development. Thus, we used whole mount in situ hybridization (WISH) to detect and localize low levels of transcript expression to developing structures and cell layers. NOGO-A gene expression was detected as early as HH3 within the primitive streak and node (as shown in HH5, Fig 3) and remained within these organizing structures until their regression. NOGO-A subsequently localized to the forming CNS during neural plate induction (HH5 Fig 3). Expression was lacking in the notochord and overlying neuroectoderm of the presumptive neural groove (HH6). Furthermore, expression was asymmetric within the primitive node, being restricted to the right side. NOGO-A expression increased during neural tube formation (HH8, Fig 3) and attained the full length of the body axis within the primitive spinal chord by HH13. Also of note was the focal presence of NOGO-A associated with forming somites (starting at HH7-8, Fig 3).

Spatiotemporal expression of NOGO-A during pivotal stages of tectal development was next characterized. The expression was then correlated with stage-specific functional markers by IHC.

Because NOGO-A expression coincides with the formation of the neural plate, we evaluated whether NOGO-A could be induced in cranial non-neural epiblasts in response to ectopic application of FGF (Fig 8A), an early initiator of neural programming. A number of FGFs are expressed during neural induction (FGF2, FGF3, FGF4, FGF8) and might play a role in the initiation of the neural program. FGF4 is often used in overexpression assays because it causes neural induction at a significantly lower concentration than other FGFs. This is believed to be due to the fact that FGF4 binds to a number of FGF receptors and is likely to reproduce the activity of other FGFs 2324. We found that NOGO-A expression was up-regulated around beads soaked in FGF4 (50 μg/ml) three hours after application in HH 3⁺/4 chick embryos (6 out of 6 embryos). This rapid FGF induced up-regulation of NOGO-A was equivalent to SOX3, an early pre-neural marker (Fig 8B). NOGO-A expression persisted around the bead for at least 9 hours (last time point examined, 5 out of 6 embryos). PBS control beads did not up-regulate either SOX3 or NOGO-A (Fig 8C,F).

When aligned with human, rat, mouse and xenopus amino acid sequences, the chicken NOGO-A sequence has a high degree of identity (>75%) in the carboxyl-terminal, reticulon-specific portion that is present in all three major isoforms (A, B and C). Multiple alignment of widely different species also revealed eight conserved regions (CR1-8) within the Nogo-A specific coding sequence (Fig 2). This conservation implies preservation of domains that are functionally significant. Previous functional domains have been identified in regions common to all Nogo isoforms [Nogo66 in the C-terminus, NiRΔ2 (rat-aa 59–172) in the shared Nogo-A/B regions], and in the Nogo-A specific region [NiGΔ20 (rat-aa 544–725)], but have been linked to inhibitory roles in neurite outgrowth and cell spreading 11. The NiGΔ20 inhibitory domain overlaps three highly conserved regions identified in our analysis (CR4-6) suggesting a conserved function across taxa. Interestingly, NiRΔ2 does not correspond to one of the highly conserved regions and previous reports have demonstrated poor preservation of this region in amphibians 28. Collectively, these data may indicate a mammalian specific role for NiRΔ2. Several highly conserved regions remain in this long protein without known functional roles. It is likely that one or more of these regions are important for Nogo-A's participation in development, however, further studies are needed to validate this prospect.

Expression of the NOGO-C isoform was broad and persistent in the developing brain by northern analysis and was in accord with adult expression profiles, where NOGO-C can be found in neurons as well as skeletal muscle 7. The larger (NOGO-A and -B) isoforms showed progressive expression with development (Fig 1C). Interestingly, we identified two NOGO-A transcripts in the chick, which confirms an earlier report identifying two NOGO-A bands by immunoblot 19. Sequence analysis against the published chicken genome verified the two bands as alternative splice products, with the smaller band resulting from a secondary splice site in exon 1, similar to what is observed in Xenopus 29. It is unclear what functional advantage these two splice variants offer, however, there appears to be differential stage-specific expression by northern analysis which suggests stage-specific roles.

The onset of neurulation is defined by formation of the neural plate. NOGO-A is induced in the neural plate during its formation and persists in structures derived from neuroectoderm. Furthermore, our data demonstrate that transformation of non-neural ectoderm into presumptive neuroectoderm by ectopic FGF is characterized by rapid induction of SOX3 (a "pre-neural" marker) and NOGO-A (Fig 8B,E). Recent microarray analysis supports this finding with the identification of NOGO-A in human and porcine neural precursor cells 30. These findings suggest that Nogo-A participates in specifying neuronal identity and/or differentiation.

However, non-neuronal NOGO-A expression was also present in the primitive streak and node, structures which precede neural plate formation (HH5, Fig 3). NOGO-A expression persists within these structures until their regression. This differential induction indicates tissue specific and temporal-spatial regulation of NOGO-A. Interestingly, there was asymmetrical right-sided NOGO-A expression within the primitive node (Hensen's node) and scant expression in the notochord and presumptive neural groove immediately overlying the notochord. This unique expression pattern is complimentary to the expression of Sonic Hedgehog (SHH), a morphogen critical to patterning of the developing neural tube (See additional file 1). NOGO-A also exhibits focal expression associated with somite formation at what may be the sites of future dorsal root ganglia and spinal nerves. We and others have shown that dorsal root ganglia have robust NOGO-A expression (See additional file 2) 1513.

The cells of the neural plate, primitive streak/node and somites all demonstrate rapid growth and patterned structural transformations. It is possible that Nogo-A may have neural-specific roles and additional broader roles related to morphogenesis. Changes in cell shape and migration are orchestrated by cytoskeletal reorganization. Nogo-A signaling has been linked to the downstream activation of the small GTPase, RhoA 31. In development, the Rho family, through Rho kinases, regulates cytoskeletal reorganization associated with changes in cell shape such as the formation of axons and dendrites 323334. Moreover, the published expression of Rho Kinase α 35 overlaps the expression of NOGO-A shown here. Inhibition of Rho kinases results in disrupted formation of the brain/neural tube, reduced caudal extension of the primitive streak and loss of left-right asymmetry. The asymmetric expression of NOGO-A within the primitive node, and its expression within the primitive streak may indicate an additional role for this protein in organogenesis and tissue identity.

Many molecules critical to CNS development have multiple functions. Molecules such as FGFs, SHH, BMPS and Wnts, once thought to play isolated roles in development, have now also been linked with several steps in neurulation and CNS development including formation of the neuronal circuitry through axon guidance and synaptogenesis 36. The expression of NOGO-A at key phases of CNS development from neural induction to definitive network establishment suggests that this factor may also have multiple functions.

To examine potential roles of Nogo-A during specific phases of CNS development we utilized the chick optic tectum. This highly structured region serves as an ideal model for understanding formation of the vertebrate brain, attaining a high level of complexity while completing most of its maturation prior to hatching. NOGO-A was present in the tectum at all stages of embryonic development observed, nevertheless, clear patterns emerged when looking at embryonic stages correlating to the specific phases of CNS development.

Proliferation is a critical part of early development of the expanding neural tube. We saw no correlation between NOGO-A expression and Proliferative Nuclear Cell Antigen (PCNA) positivity when comparing neuroepithelium during peak proliferative and relative quiescent stages (E5 vs E10, respectively) 37. Thus, it is unlikely that Nogo-A has a role in proliferation.

From E6 onward a migratory zone of post-mitotic neurons can be visualized in the tectum 38. NOGO-A expression in the E7 tectum was homogenously expressed across the pre-migratory generative zone, the expanding migratory layer, and the post-migratory, first neuronal lamina. Immunostaining for neuronal and glial cells linked NOGO-A expression to the soma of neurons, however, NOGO-A expression was not limited to the neuronal population actively undergoing migration. Thus, our data does not support a role for Nogo-A limited to directing the migration of neurons. Rather, the continued expression of NOGO-A within all populations of developing neurons is further support of a role marking early neural identity and may be related to the state of maturity/differentiation. This concept is in agreement with previous studies that reported a high neuronal NOGO-A mRNA expression during differentiation of the spinal cord 15, and a down regulation of NOGO-A mRNA following migration and terminal differentiation of rat olfactory neurons 25 and cerebellar granule cells 17.

Neuronal differentiation is also characterized by the outgrowth of neurites and the formation of synapses. Neuritogenesis and synaptogenesis coupled with later refinement of connections by cell death and neurite pruning comprise the process of "network establishment." The Stratum Griseum Centrale (SGC) of the chick optic tectum is sparsely populated by large, multi-polar principal efferent neurons with extensively branched dendrites and robust projecting axons 39. By E10, extension of axons and dendrites is abundant in the SGC and intense expression of NOGO-A can be seen within these large neurons. Heightened NOGO-A expression can also be seen in nearby tectal associated nuclei, which are conspicuous by their large, projecting neurons.

By E14, the first synaptic junctions of the tectum are observed. However, their numbers rise drastically between E18 and the first hours after hatching with more intense retinal input 40. Loss of retinal input to the tectum can be disturbed by embryonic enucleation of the chick eye optic anlagen. Deafferentation causes neuronal degeneration of the SGC through loss of synaptic input and prevention of forming synaptic connections 41. Correspondingly, there is an overall decrease in NOGO-A expression within the SGC and the tectal associated nuclei suggesting NOGO-A expression is dependent upon synaptic activity (Caltharp, unpublished data).

Increased Nogo-A immunoreactivity has been demonstrated at the onset of axon growth in developing rat olfactory neurons 25 and also within sprouting dendrites of cerebellar Purkinje cells 17. Interestingly, over expression of Nogo-A within COS cells leads to the formation of long processes that resemble neurites 42. As mentioned earlier, Nogo-A can regulate RhoA dependent cytoskeletal reorganization which is intensely active during neuronal differentiation and neuritogenesis. Collectively, these data combined with our findings provide strong support for a Nogo-A specific role in neuronal differentiation and neuritogenesis.

At E10, NOGO-A expression within the tectum is strictly neuronal. Oligodendrocyte related myelination in the chick tectum takes place between E12–E17 22 and accordingly, unmyelinated fiber tracts within the tectum are distinctive for an absence of NOGO-A. By E20, these same fiber tracts have begun to acquire myelination by oligodendrocytes and are now positive for NOGO-A expression. Studies marking glial differentiation found Nogo-A to be expressed at all stages of oligodendrocyte development 1813 and our data found NOGO-A expression within cells of tectal associated nuclei that were neither positive for neuron, astrocyte or glial progenitor specific stains. This later NOGO-A expression within smaller cells of the tectum is a likely sign of increased oligodendrocyte formation and may reflect the beginning transition of expression from neuronal differentiation and network establishment to myelin derived maintenance of the circuitry.

Within a region specific for Nogo-A, amino acid sequences conserved between human and rat were identified and used to generate a 5' degenerate primer (5'-GCCTGAAGGYCTGACKCC-3') and a 3'degenerate primer (5'-GGTGCTTCCATAACAATRTCAGGC-3'). These primers were used to isolate a 219 bp fragment of chicken NOGO-A [Genbank: AY494005] from embryonic chick brain cDNA (E10). This fragment was cloned into plasmid vector pCRII-Topo (Invitrogen) with the capacity for forward and reverse transcription. 3' extension of the initial isolated fragment was performed using the designed 5'p gene specific primer and an Oligo-dT primer as provided in the GeneRacer Kit (Invitrogen) for amplification, cloning and sequencing of the cDNA. A probe for the common region of the NOGO gene was based on sequencing results of 3' NOGO-A using the common 5' primer (5'-gttgttgacctcctttactgg-3') and common 3' primer (5'-ccagtatcagtaatgtcagacc-3'). The resulting fragment was cloned as described above.

For complete 5' sequencing of NOGO-A, we constructed primers based on the chicken genome (Gallus_gallus 1.043) to regions that appeared to match start sites of the gene within other species. Triple Master (Eppendorf) GC rich PCR protocol with 4% DMSO was used for sequence isolation [Genbank: AY843529]. Protein multiple alignment was performed using Clustal X (version 1.83) as described 44. Regions of significant conservation were determined by receiving a score of 70% sequence similarity based on amino acid identity (as indicated by asterisk) and/or conservation of one of the 'strong' groups (as indicated by colon), see 44. The alignment was run on a downloaded program 45.

White Leghorn fertilized eggs were obtained from Hyline International (Lakeview, CA). Eggs were incubated at 39°C in a humidified chamber, and embryonic age was determined according to Hamburger and Hamilton's (HH) staging system 46. Chicks representing stages 3–11,15, 17, 25, 30, 33, 36, and 45 were isolated for study.

Chicks at stages 3 to 11 (14 hr to 45 hr incubation) were harvested with a filter paper ring support, rinsed in cold PBS and fixed in MEMFA pH 9 (0.1 M MOPS pH7.4; 2 mM EGTA; 1 mM MgSO₄; 3.7% formaldehyde) overnight at 4°C. Embryos were then briefly rinsed in PBS, transferred to ice cold 90% methanol and stored at -20°C until processed. Anti-sense riboprobe from the 219 bp NOGO-A fragment was generated from linearized plasmid using digoxigenin-tagged rUTP (Roche) in the substrate mix following standard whole mount in situ protocol 47. Plasmids for SOX2 and SOX3 probes were kind gifts from Drs. Domingos Henrique, R. Lovell-Badge and P. Scotting. Subsequent steps were carried out either manually for the earliest stages (3–11) or with the InsituPro automated whole mount ISH (Intavis AG) for stages 15 and 17. Embryos were rehydrated and treated with proteinase K (10 μg/ml) for 5 min. Probe was applied overnight (~12 hr) at 58°C and post hybridization washes were carried out at 63°C. After pre-blocking embryos with 2% blocking reagent (Roche), a high-affinity anti-digoxigenin antibody conjugated to alkaline phosphatase (Roche) was applied. Colorization reaction with BCIP-NBT in 12.5% PVA was performed in the dark and checked periodically under a dissecting microscope until reaction was deemed complete (approximately 2–3 hr).

Whole embryos of chicks at stages 15 and 25 (E 2.5 and 4.5, respectively) and dissected brains from chicks at stages 30, 33, 35 and 45 (E 7,8, 10 and 20, respectively) were harvested, fixed in 4% paraformaldehyde (PFA) for 24 hours at 4°C, dehydrated in a series of ethanol and xylene gradients, infiltrated with, and embedded in paraffin. In situ hybridization was performed with ³⁵S-labeled riboprobes on 5 μm sections for normal developing brain as previously described 47 with a hybridization and wash temperature of 58°C and 63°C, respectively. Sense probes were also generated and hybridized in a similar manner to document anti-sense probe specificity (data not shown). Following autoradiography (Kodak BioMax MR film), the slides were counter-stained with Hoechst 33258 dye (2 mg/ml) to highlight nuclei and then visualized and digitally recorded (Sony DKC-5000) on a fluorescence compound microscope. We used darkfield illumination for specific expression and DAPI filtered-fluorescence for nuclear localization and general morphology. The two images were then overlaid and pseudo-colored in Adobe Photoshop 6.0.

Total RNA from stage HH 15, 25, 36 and 45 (E2.5, 5, 10, 20, respectively) chick brain was isolated using RNA BEE (Tel-Test) as per the manufacturer's protocol. PolyA RNA was extracted from the total RNA using Oligotex Spin Column mRNA kit (Qiagen). Two micrograms of Millenium Marker-F (Ambion) was run with the mRNA samples to indicate relative band size. The RNA samples and 1% formaldehyde gel were prepared as suggested in the Qiagen Oligotex Handbook Appendix G and transferred to a nitrocellulose membrane in alkaline conditions 48. A ³²P dCTP labeled probe of the 219 bp, common NOGO or β-actin fragment was prepared using the High Prime labeling kit (Roche). Hybridization of the labeled probe was performed at 68°C for 1 hour in QuickHyb Hybridization Solution (Stratagene) as per the manufacturer's protocol. The membrane was exposed to Kodak BioMax MR film at -80°C prior to developing. Controls for mRNA were performed using a PCR generated chick β-actin plasmid as described 49. Densitometry readings of the autoradiographic film were taken using a Kodak MultiImage Light Cabinet and assessed on ChemiImager 4400 v5.1 software program using spot densitometry.

Sections adjacent to those processed for ISH were de-paraffinized and subjected to antigen retrieval (Citra Plus, BioGenex Microwave Antigen Retrieval System). Immunostaining was performed by the BioGenex i6000 (Model 1.0) automated staining system using standard immuno-protocols. Briefly, after blocking (3% H₂O₂ in 10% methanol and Powerblock, BioGenex), primary antibodies were applied to the slides for 40 minutes using Vimentin (1/200, Ventana), Neurofilament M (1/3200, Chemicon), GFAP (8 mg/ml, Chemicon), and PCNA (1/200, Dako). A multi-species, biotinylated, secondary antibody (Multi-link, Biogenex) was then applied to the slides for 20 minutes. The detection signal was amplified with horseradish peroxidase-conjugated streptavidin (Superlabel, BioGenex) and then visualized using a permanent (non-aqueous) chromogen, Romulin AEC (Biocare). Positive and negative (BSA) controls were performed for each staining run (Data not shown).

Heparin acrylic beads (Sigma) of 80–120 μm size were manually isolated using a stereomicroscope, washed multiple times in PBS, soaked in 50 μg/ml FGF-4 (R&D Systems) at 4°C for 1–2 hr and then washed in PBS prior to implant. HH4 chick embryos were cultured using the EC culture method 50. Using a 0.01 mm sharpened tungsten needle beads were pushed between endoderm and ectodermal layers of ventral side up embryos and placed in anterior, non-neural ectoderm 51. Embryos were then harvested in MEMFA at 3, 6, or 9 hours post implant and processed for whole mount in situ hybridization as described above.