SEARCH

SEARCH BY CITATION

Keywords:

  • brain;
  • spinal cord;
  • neuronal migration;
  • axon guidance;
  • actin cytoskeleton

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED

We investigated the expression of the three known Slit-Robo GTPase activating protein (srGAP) genes in the developing murine nervous system using in situ hybridization. The three genes are expressed during embryonic and early postnatal development in the murine nervous system, showing a distinct pattern of expression in the olfactory system, the eye, forebrain and midbrain structures, the cerebellum, the spinal cord, and dorsal root ganglia, which we discuss in relation to Slit-Robo expression patterns and signaling pathways. We also report srGAP2 expression in zones of neuronal differentiation and srGAP3 in ventricular zones of neurogenesis in many different tissues of the central nervous system (CNS). Compared to srGAP2 and srGAP3, the onset of srGAP1 expression is later in most CNS tissues. We propose that these differences in expression point to functional differences between these three genes in the development of neural tissues. J. Comp. Neurol. 513:224–236, 2009. © 2009 Wiley-Liss, Inc.

Brain development is dependent on the finely orchestrated migration of numerous populations of neuronal precursors from proliferative ventricular and subventricular zones to specific target sites for differentiation as well as the guided projection of axons toward their target neurons for synapse formation. Guided migration and axon projection rely on attractive and repulsive guidance cues, such as the Slit ligand and its receptor Robo, which were originally found to be necessary for axon guidance in Drosophila (Kidd et al.,1999). We now know that Slit-Robo signaling has conserved roles in development of the vertebrate nervous system (Brose et al.,1999). In particular, Slit-Robo signaling is important for normal midline crossing by commissural axons in the spinal cord (Long et al.,2004; Sabatier et al.,2004; Chen et al.,2008) and neuronal precursor migration (Hu,1999; Wu et al.,1999; Marin et al.,2003; Nguyen-Ba-Charvet et al.,2004) as well as in axon branching (Wang et al.,1999; Ma and Tessier-Lavigne,2007).

Neuronal precursor migration and axon extension is dependent on cytoskeletal reorganization, which must be tightly controlled for directional migration and axon guidance. The family of Rho-GTPases, including Rac, Cdc42, and Rho, have emerged as key regulators of cytoskeletal dynamics (Hall,1998) and normal neuronal development (Newey et al.,2004) and have been shown to act downstream of axon guidance cues to control cytoskeletal reorganization for growth cone turning (Huber et al.,2003). Rho-GTPases cycle between an active (GTP-bound) conformation and an inactive (GDP-bound) conformation. Guanine nucleotide exchange factors (GEFs) and GTPase activating proteins (GAPs) act antagonistically with each other to tightly regulate Rho-GTPase function; GEFs catalyze nucleotide exchange and mediate activation, while GAPs promote GTP hydrolysis, leading to inactivation.

A study by Wong et al. (2001) identified a family of GAP proteins which bind to the CC3 cytoplasmic domain of Robo1, thus linking the guidance factors Slit and Robo with the actin cytoskeleton. These proteins were named the Slit-Robo GTPase activating proteins (srGAPs) and consist of srGAP1, srGAP2, and srGAP3, which have a highly conserved structure, all containing an FCH domain, a GAP domain, and an SH3 domain, which binds to Robo1 (Wong et al.,2001). The GAP activity of the srGAP proteins appears to be specific to certain RhoGTPases: srGAP1 predominantly regulates Cdc42 signaling (Wong et al.,2001), while srGAP3 has been shown to preferentially downregulate Rac1 activity (Endris et al.,2002; Soderling et al.,2002). The specificity of srGAP2 GAP activity is currently undefined.

Since the identification of the srGAP family, we have learned that they have some important functions in neuronal development. srGAP1 is required for Slit-mediated repulsion of migrating subventricular zone forebrain neurons through a downregulation of Cdc42 activity in vitro (Wong et al.,2001), while functional disruption of the srGAP3/MEGAP protein is associated with severe mental retardation (Endris et al.,2002), implicating srGAP3/MEGAP in neuronal development for normal cognitive function. Indeed, srGAP3 has been shown to regulate neuronal development and synaptic plasticity in a complex with WAVE1 (Soderling et al.,2007). Despite these interesting studies, our understanding of the role of the srGAP genes is in its infancy.

In an effort to further understand the roles of these three genes in the development of the nervous system, we have performed extensive expression analysis in the developing murine central nervous system (CNS) using in situ hybridization. Previous studies have described limited expression patterns for srGAP3 (Waltereit et al.,2008) and for srGAP1 and srGAP2 (Wong et al.,2001) mRNA during development, but this is the first study to closely examine and compare srGAP gene expression in multiple CNS tissues throughout embryonic and early postnatal development. We have uncovered key differences in the expression patterns between the three srGAP genes, which we believe provide useful insights into possible functional differences. We also discuss the significance of our findings in relation to Slit-Robo signaling.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED

Animals

NMRI mice (Charles River Laboratories, Germany) were analyzed at embryonic (E) and postnatal (P) stages E11.5, E12.5, E13.5, E14.5, E16.5, P1, and P7. The day of the vaginal plug was taken as day E0.5 and the day of birth was taken as day P0. Pregnant females were sacrificed by cervical dislocation in accordance with a previously approved institute protocol.

Embryonic and postnatal tissues were fixed by immersion in 4% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.4, at 4°C. The fixed tissues were then rinsed in PBS and cryopreserved by immersion in 30% sucrose in PBS at 4°C. Tissues were then embedded in Jung tissue freezing medium (Leica Microsystems, Nussloch, Germany) compound in a dry ice and ethanol bath and stored at −80°C until sectioning.

Cloning of srGAP1, 2, and 3 cDNAs

Mouse cDNA was used as a template in polymerase chain reactions (PCRs) for all three probes. For the srGAP1 probe a 678 bp PCR product was amplified using forward primer 5′-CATGAATTCTTGGAGGAGGTGGACCAAGACGC-3′ and reverse primer 5′-GATCTCGAGTTACTGCCCTGATGATGTG- GACC-3′. For the srGAP2 probe we generated a 675-bp PCR product using forward primer 5′-GATGAATTCGCTGAAGAC- TCCACCCAGGACGT-3′ and reverse primer 5′-GATGAATTC- GCTGAAGACTCCACCCAGGAC GT-3′. These PCR products were cloned into a pBSII vector (Stratagene, La Jolla, CA) using EcoRI and XhoI. For the srGAP3 probe, a 534-bp PCR product was generated using forward primer 5′-GATGGGAATTCCCAGGACATGGATGATGCCTTC-3′ and reverse primer 5′-CTAGATACTCGAGGTGCTCATGGTCTTC- TCGATG-3′ and the resulting PCR product was cloned into the pST-blue Acceptor Vector (Novagen, Madison, WI).

Riboprobe synthesis

In vitro transcription of the probes was carried out using the Megascript T3 (srGAP1 sense, srGAP2 sense), T7 (srGAP1 antisense, srGAP2 antisense, srGAP3 sense), or SP6 (srGAP3 antisense) High Yield Transcription Kit (Ambion, Austin, TX) and probes were labeled with digoxigenin-UTPs (Roche Diagnostics, Mannheim, Germany).

In situ hybridization with digoxigenin-labeled riboprobes

Embedded tissue was sectioned using a cryostat (Leica Microsystems, Nussloch, Germany). Tissue sections were hybridized with digoxigenin-labeled antisense and sense probes overnight with hybridization buffer (10 mM Tris, 180 mM NaCl, 6 mM NaH2/PO4, 5 mM Na2/HPO4, 5 mM EDTA, 50% formamide, 10% dextransulfate, 0.5× Denhard's, pH 7.4). Hybridization was carried out at 68°C in a humidified box (humidified with 50% formamide, 1× SSC). After hybridization the samples were washed three times at 68°C in wash buffer (50% formamide, 1× SSC, 0.2% Tween 20 in DEPC-H2O), then rinsed twice for 30 minutes in MAB-T buffer (100 mM maleic acid, 150 mM NaCl, 0.1% Tween 20, pH 7.5). Samples were then incubated for 30 minutes in blocking solution (20% horse serum in MAB-T) before incubating overnight with antidigoxigenin antibody (Roche Diagnostics) at a concentration of 1:5,000 in blocking solution. Tissue sections were then rinsed four times in MAB-T buffer, followed by twice in AP buffer (100 mM Tris/HCl, pH 9.5, 100 mM MgCl2, 0.1% Tween 20, 0.25 mg/mL Levamisol). Afterwards, samples were incubated in NBT/BCIP color reagent (Roche Diagnostics), diluted to 1:100 in AP buffer. Following the color reaction, tissue sections were washed three times in PBS before mounting. Imaging was carried out on a Nikon 90i upright microscope with a Nikon DXM 1200C color camera. Optimization of brightness and contrast of the images was performed using Adobe Photoshop software (San Jose, CA). Each in situ reaction was repeated three times and three different animals were sectioned for each developmental stage to eliminate any variability in expression between animals. Brain structures were identified using the Chemoarchitectonic Atlas of the Developing Mouse Brain (Jacobwitz and Abbott1998).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED

Development of the cerebral cortex

The cerebral cortex develops from the wall of the telencephalic vesicle and is composed of six distinct layers. These layers are formed as cortical neurons migrate to their final positions between E11 and E18 in the mouse brain (Gupta et al.,2002). At E11.5 the neocortex is composed of the ventricular zone containing proliferative neuronal precursors and the preplate, which is a product of radial migration of postmitotic neurons from the ventricular zone. At this stage, only srGAP2 and srGAP3 genes were expressed in the telencephalon, with distinct localization patterns. srGAP2 was restricted to the preplate (PP; Fig. 1B), while srGAP3 was detected throughout the PP and the ventricular zone (VZ; Fig. 1C).

thumbnail image

Figure 1. srGAP1, srGAP2, and srGAP3 mRNA expression during development of the cerebral cortex from embryonic day (E) 11.5 until postnatal day (P) 7. A: srGAP1 mRNA was not detected in the telencephalon at E11.5 (arrow). B: srGAP2 mRNA was detected in the preplate (PP) at E11.5 but not the ventricular zone (VZ). srGAP2 mRNA expression was also found in the trigeminal ganglion (TG), the optic pit (OP), olfactory pit (Olf), and the superior colliculus differentiating zone (scdz). C: srGAP3 mRNA was expressed in the PP and VZ at E11.5. Expression was also detected in the TG, OP, Olf as well as neuroepithelium of the superior colliculus (scn) D: At E13.5, srGAP1 mRNA was detected in the cortical plate (CxP) of the cerebral cortex and very weakly throughout the subplate (SP) and intermediate zone (IZ). E: srGAP2 mRNA was detected primarily in the CxP of the developing cortex at E13.5 and very weakly in the IZ. F: srGAP3 mRNA was detected in the CxP and more weakly in the SP and IZ. G: srGAP1 mRNA was detected primarily in the CxP at E16.5 and also in IZ and VZ. H: srGAP2 mRNA was distinctly observed in the CxP, IZ, and VZ but not in the SP. I: srGAP3 mRNA was detected in the CxP, IZ, and VZ of the developing cortex at E16.5. J: At P7 srGAP1 mRNA localized primarily to cortical layers II to V, with some scattered expression in layer VI. Layer I was free of srGAP1 mRNA. K: srGAP2 mRNA was expressed in cortical layers II to IV at P7, with strongest expression in the upper layers. Expression was also detected in layer V and more diffusely throughout layer VI. No expression was detected in layer I. L: srGAP3 mRNA was detected in all layers of the cortex at P7, except layer I, which clearly expressed no srGAP3 mRNA. A–C represent sagittal sections; scale bars = 1,000 μm. D–F represent sagittal sections; scale bars = 200 μm. G–I represent sagittal sections; scale bars = 500 μm. J–L represent coronal sections; scale bars = 1,000 μm.

Download figure to PowerPoint

At E13.5 a second population of postmitotic neurons has migrated through the intermediate zone to split the PP into the marginal zone and the subplate, with the cortical plate separating the two. At this stage we detected the first expression of srGAP1 mRNA in the developing cortex, which was restricted to the cortical plate (CxP; Fig. 1D). srGAP2 mRNA was also restricted to the CxP (Fig. 1E). At E13.5, srGAP3 mRNA was also most prominently expressed in the CxP, with some weaker expression in the subplate (SP), intermediate zone (IZ), and VZ (Fig. 1F).

At E16.5 the thickness of the CxP increased, due to the continual radial migration of neuronal precursors from the VZ and all three srGAP genes were strongly expressed in the CxP at this stage (Fig. 1G–I). srGAP1, srGAP2, and srGAP3 mRNAs were also detected in the IZ and VZ at E16.5 (Fig. 1G–I), although srGAP1 expression was very weak in these layers.

By P7 it is possible to distinguish the six layers of the cortex in tissue sections. None of the srGAP genes were expressed in layer I at this stage (Fig. 1J–L; CI). srGAP1 mRNA was most concentrated in cortical layers II – V, with more diffuse expression in layer VI (Fig. 1J). srGAP2 mRNA localized to layers II–V and more diffusely in layer VI (Fig. 1K), while srGAP3 was expressed in all the CxP except layer I, but appeared to be most concentrated in layers II–IV (Fig. 1L).

srGAP expression in the thalamus

The neocortex and the thalamus are linked through a reciprocal connection, which forms between E13 and E18 in mice. At E13.5, srGAP1 mRNA was detected in the dorsal thalamus (dTh; Fig. 2A). srGAP2 mRNA was strongly detected in the dTh and more weakly in the underlying ventral thalamus (vTh; Fig. 2B). We also detected srGAP3 mRNA in the dTh and vTh at E13.5 (Fig. 2C). At P1, all srGAP genes were detected in the dTh and vTh (Fig. 2D–F) and srGAP3 was most strongly expressed in the vTh (Fig. 2F). At P7, all three srGAP genes were expressed in the dorsal lateral geniculate nucleus (DLG; Fig. 2G–I). srGAP1 and srGAP3 were also expressed in the ventral lateral geniculate nucleus (VLG; Fig. 2G,I). We also detected srGAP1 and srGAP2 mRNA expression in the ventrolateral thalamic nucleus at P7, but not srGAP3. In the ventral posterolateral nucleus, srGAP1 and srGAP3 expression was detected at P7, but no srGAP2. No staining was detected in sections labeled with sense probes (data not shown).

thumbnail image

Figure 2. srGAP mRNA expression in the developing thalamic and hippocampal region from embryonic day (E) 13.5 until postnatal day (P) 7. A: srGAP1 mRNA was detected in the dorsal thalamus (dTh) at E13.5. B: The dTh strongly expressed srGAP2 at E13.5 and expression was also detected in the ventral thalamus (vTh). C: srGAP3 mRNA was detected in the vTh and the dTh at E13.5. Expression of srGAP3 mRNA was also strongly detected in the hippocampal neuroepithelium (H) at E13.5. D: srGAP1 mRNA was detected in the developing hippocampus at P1. Expression was strongest in the CA3 region, mRNA was more weakly detected in the CA1 region and even less so in the dentate gyrus (DG). Weak expression was also detected in the dTh and the vTh. E: At P1, srGAP2 mRNA was detected in the CA3 region of the developing hippocampus. Expression was weakly detected in the CA1 region but was absent from the DG. srGAP2 expression was also weakly detected in the vTh and the dTh. F: srGAP3 mRNA was detected throughout the developing hippocampus and was abundant in both the CA1 and CA3 regions. Expression was also observed in the DG. srGAP3 mRNA was also detected in the vTh and more weakly in the dTh at P1. GI: G: At P7, srGAP1 mRNA was detected throughout the developing hippocampus in the CA1 region, CA3 region, and DG. Expression was also detected in the dorsal lateral geniculate nucleus (DLG) and more weakly in the ventral lateral geniculate nucleus (VLG). H: srGAP2 expression was also detected in the hippocampus at P7, although expression was more or less restricted to the CA1 region. Weak expression was observed in the CA3 region, but was absent from the DG. Expression was also detected in the DLG. I: srGAP3 mRNA was expressed throughout the hippocampus at P7 and was detected in the CA1 region, the CA3 region and the DG. At P7, expression was also observed in the DLG and VLG. Cx, cerebral cortex. All sections are sagittal. Scale bars = 500 μm in A–F; 1,000 μm in G–I.

Download figure to PowerPoint

srGAP expression during embryonic and postnatal hippocampal development

In order to gain more insight into the role of srGAP genes in hippocampal development, we investigated their expression in the embryonic and early postnatal hippocampus.

At E13.5 only srGAP3 mRNA was detected in the developing hippocampus (H; Fig. 2C). At P1, srGAP1 mRNA was most strongly expressed in the CA3 region of the H (Fig. 2D) but was also detected in the CA1 region and the dentate gyrus (DG; Fig. 2D). srGAP2 was only very weakly detected in the CA1 region at P1 (Fig. 2E). Stronger expression was observed in the CA3 region (Fig. 2E), while the DG was free of srGAP2 mRNA at P1 (Fig. 2E). srGAP3 was expressed throughout the H but was most concentrated in the CA3 region at P1 (Fig. 2F). At P7, srGAP1 was detected in the CA1 and CA3 regions and the DG (Fig. 3G). srGAP2 mRNA was still absent from the DG at P7 (Fig. 2H) and was only very weakly detected in the CA3 region (Fig. 2H). srGAP2 expression was strongest in the CA1 region at this stage (Fig. 2H). srGAP3, like srGAP1, was expressed throughout the H at P7, in the CA1 and CA3 regions and the DG (Fig. 2I).

thumbnail image

Figure 3. srGAP1, srGAP2, and srGAP3 mRNA expression in the developing cerebellum from embryonic day (E) 13.5 until postnatal day (P) 7. A: At E13.5, srGAP1 mRNA is weakly detected in the roof of the midbrain in the superior colliculus differentiating zone (scdz), but is largely absent from the superior colliculus neuroepithelium (scn). At this stage, no srGAP1 mRNA was detected in the developing cerebellar primordium (CbP). A′: Coronal section taken through the CbP, as indicated by the line in panel A. No srGAP1 expression was detected in the granule cells at this stage (arrows). B: srGAP2 mRNA was strongly detected in the roof of the midbrain, particularly in the scdz and was absent from the scn. Expression was also detected in the cerebellar plate (CP) of the developing CbP and in the cerebellar nuclei differentiating zone (asterisk). B′: Coronal section taken through the CbP at the position indicated by the line in B. Expression was not detected in the granule cells at this stage (arrows). C: srGAP3 mRNA was detected throughout the roof of the midbrain at E13.5, but was particularly concentrated in the scn. Expression was also detected throughout the inferior colliculus (IC). srGAP3 mRNA was also detected in the CP and in the cerebellar nuclei differentiating zone (asterisk) at this stage. C′: srGAP3 mRNA was detected in the granule cells in coronal sections of the CbP at E13.5 (arrows). The coronal section was cut at the position indicated by the line in C. D: At E16.5, srGAP1 mRNA was weakly detected in the superior colliculus (SC) of the midbrain, the IC, and the CP. E: srGAP2 mRNA was detected in the scdz at E16.5. It was also strongly detected in the inferior colliculus differentiating zone (icdz). Expression was also detected throughout the CP. F: At E16.5, srGAP3 mRNA was detected in scn and throughout the icn and icdz. G: srGAP1 mRNA was expressed in the external granule layer (EGL), Purkinje cell layer (PL), and internal granule layer (GL) at P1. Expression was also detected in a region of the inferior colliculus (arrows). H: srGAP2 expression in the cerebellum was restricted to the PL at P1. srGAP2 expression was also detected in a region of the inferior colliculus at P1 (arrows). I: srGAP3 mRNA was detected primarily in the PL of the cerebellum at P1 and more weakly in the EGL and GL at this stage. Expression was also detected in the scn at P1 (arrows) and in a region of the inferior colliculus (IC). The boxed regions in G–I are shown at higher magnification in the top right-hand corner. J: At P7, srGAP1 expression was strongly expressed in the EGL, PL, and GL of the developing cerebellum. K: srGAP2 mRNA was restricted to the PL in the developing cerebellum at P7. L: At P7, srGAP3 mRNA was observed throughout the developing cerebellum, in the EGL, PL, and GL. Expression was also detected in a population of deep cerebellar nuclei (CN). Insert images represent a magnified view of the boxed regions in J–L. icn, inferior colliculus neuroepithelium; icdz, inferior colliculus differentiating zone; MO, medulla oblongata; IC, inferior colliculus; CPx, choroid plexus. All sections are sagittal, except A′–C′, which are coronal. All scale bars = 500 μm.

Download figure to PowerPoint

srGAP expression in the superior and inferior colliculi

The superior and inferior colliculi are formed from the dorsal portion of the mesencephalon and neurogenesis in the superior colliculus peaks at E13.5. At E11.5, srGAP1 mRNA was weakly detected in the superior colliculus differentiating zone (scdz; Fig. 1A). This continued until E13.5 (Fig. 3A). By E16.5, expression was detected throughout the superior colliculus (SC) rather than restricted to the differentiating zone and detected for the first time in the inferior colliculus (IC; Fig. 3D). srGAP1 continued to be expressed in a specific region of the IC at P1 (Fig. 3G; arrows).

At E11.5, srGAP2 was strongly detected in the scdz (Fig. 1B) and was absent from the superior colliculus neuroepithelium (scn). This expression pattern was maintained at E13.5 (Fig. 3B) up until E16.5 (Fig. 3E), with no expression detected in the scn. Expression in the IC was also restricted to the differentiating zones at E13.5 (icdz; Fig. 3B) and E16.5 (Fig. 3E). At P1, weak expression of srGAP2 was detected in a region of the IC (Fig. 3H; arrows).

srGAP3 mRNA was detected throughout the SC and IC structures at the analyzed stages, but expression was most concentrated in the neuroepithelial layers until E13.5, the peak of neurogenesis. At E11.5, srGAP3 was detected most strongly in the scn (Fig. 1C) and this was also true at E13.5 (Fig. 3C) and at E16.5 (Fig. 3F). At E13.5 and E16.5, srGAP3 expression was detected throughout the IC (icn, icdz; Fig. 3C,F). By P1, srGAP3 was still detected in the scn (Fig. 3I; arrows) and in a region of the IC (Fig. 3I).

srGAP expression in the developing cerebellum

srGAP1 was not expressed in the developing cerebellar primordium at E13.5 (CbP; Fig. 3A). We first detected srGAP1 mRNA in the developing cerebellum at E14.5 (data not shown, see Table 1), where it localized to the base of the CbP, presumably a population of migrating Purkinje cells forming the cerebellar plate. The granule cells were free of srGAP1 expression at E13.5 (Fig. 3A′; arrows). At E16.5, srGAP1 mRNA was expressed more diffusely throughout the CbP, as Purkinje cells migrate radially from their zone of mitosis, forming the cerebellar plate (CP; Fig. 3D). By P1, srGAP1 was detected in the Purkinje layer (PL; Fig. 3G), which is not completely defined at this stage. srGAP1 mRNA was also detected in the external granule layer (EGL; Fig. 3G) and in some granule cells forming the granule layer (GL) deep to the PL (Fig. 3G). By P7, srGAP1 was expressed in the EGL, PL, and GL layers of the developing cerebellum (Fig. 3J). The molecular layer was free of staining.

Table 1. Summary of srGAP Gene Expression in the Developing Murine Nervous System
 E11.5E12.5E13.5E14.5E16.5P1P7
  1. -, not expressed; nd, not defined; E, embryonic day; P, postnatal day.

Telencephalon       
Eye       
Neuroblastic layer- - 3- - 3- - 31 - 31 - 3ndnd
Retinal ganglion layer- 2 3- 2 3- 2 31 2 31 2 3ndnd
Olfactory system       
Olfactory pit- 2 3      
Mitral cell layer  1 2 31 2 31 2 31 2 3nd
Granule cell layer  - - 31 - 31 - 31 - 3nd
Olfactory epithelium - 2 -- 2 -- 2 -- 2 -- 2 -nd
Vomeronasal organ - 2 3- 2 31 2 3- 2 3nd
Cerebral cortex       
Ventricular zone- - 3- - 31 2 3
Preplate- 2 3- 2 3     
Marginal zone    
Subplate  1 - 3  
Cortical plate  1 2 31 2 31 2 3  
Intermediate zone  1 2 31 - 31 2 3  
C I     
C II – IV     1 2 31 2 3
C V     1 2 31 2 3
C VI     1 2 31 2 3
Hippocampus       
Hippocampal neuroepithelium - - 3- - 31 2 31 2 3  
CA1 region     1 2 31 2 3
CA3 region     1 2 31 2 3
Dentate gyrus     1 - 31 - 3
Mesencephalon, rhombencephalon and spinal cord       
Thalamus       
Dorsal lateral geniculate nucleus 1 2 31 2 31 2 -1 2 31 2 31 2 -
Ventral lateral geniculate nucleus - 2 31 2 31 2 -1 - 31 2 31 - 3
Ventrolateral nucleus ndndndnd1 2 -
Ventral posterolateral nucleus ndndndnd1 - 3
Superior colliculus       
Neuroepithelium- - 3- - 3- - 3- - 3- 2 3- - 3nd
Differentiating zone- 2 3- 2 31 2 31 2 -1 2 -nd
Inferior colliculus       
Neuroepithelium- - 3- - 3- - 3- - 3- - 31 2 3nd
Differentiating zone- 2 3- 2 3- 2 3- 2 3- 2 3nd
Cerebellum       
Cerebellar nuclei differentiating zone 1 2 3- 2 3  
Cerebellar plate - 2 3- 2 3- 2 31 2 3  
External granule layer nd- - 3ndnd1 - 31 - 3
Molecular layer     
Purkinje layer     1 2 31 2 3
Granule layer     1 - 31 - 3
Deep cerebellar nuclei      - - 3
Spinal cord       
Dorsal horn- 2 31 2 31 2 3ndndndnd
Ventral horn1 2 31 2 31 2 3ndndndnd
Floor plate- 2 3- 2 3- 2 3ndndndnd
Roof plate- 2 3- 2 3- 2 3ndndndnd
Dorsal root ganglia- 2 3- 2 3- 2 3ndndndNd

srGAP2 expression was restricted to Purkinje cell populations during development of the cerebellum (Fig. 3B,E,H,K). At E13.5, when Purkinje cells undergo their final mitotic divisions and begin to migrate radially (Miale and Sidman,1961), srGAP2 is strongly expressed at the base of the CbP, in a region we assume to be the CP, an irregular layer formed by migrating Purkinje neurons (Fig. 3B). Deep cerebellar neurons are the targets of Purkinje cells and a population of differentiating deep cerebellar nuclei expressed srGAP2 at this stage (Fig. 3B; asterisk). srGAP2 was not detected in the granule cells at E13.5 (Fig. 3B′; arrows). At E16.5, srGAP2 expression in the CbP is more diffuse, corresponding to the radial migration of Purkinje cells and growth of the CP (Fig. 3E). At P1, srGAP2 localizes to the developing PL (Fig. 3H) and by P7, srGAP2 expression strongly defines the undulating PL (Fig. 3K).

At E13.5, srGAP3 was expressed in the CP as Purkinje cells begin to migrate (Fig. 3C) and in a population of differentiating deep cerebellar nuclei (Fig. 3C; asterisk). Expression was also detected in the granule cells at E13.5 (Fig. 3C′; arrows), which originate in the rhombic lip (not identified) (Wingate and Hatten,1999). At E14.5, srGAP3 mRNA was more weakly detected in the CP (data not shown), presumably as Purkinje cells migrated radially and expression became more diffuse. At E16.5 we detected only a very faint staining in the CP (Fig. 3F). At P1, srGAP3 mRNA was detected in the developing PL (Fig. 3I) and only very weakly in the EGL and GL (Fig. 3I). By P7, srGAP3 was expressed in the EGL (Fig. 3I), the PL (Fig. 3I), and the GL (Fig. 3I). Expression was also detected in a population of deep cerebellar nuclei (CN; Fig. 3I). The molecular layer was free of staining.

srGAP expression in the embryonic spinal cord

Spinal cord commissural neurons extend their axons dorsoventrally and begin to cross the midline at the floor plate at E11.5. At E11.5, srGAP1 mRNA was expressed specifically in the ventral horns of the spinal cord (VH; Fig. 4A) and was absent from the dorsal root ganglia (DRGs). At the onset of midline crossing at E11.5, srGAP2 mRNA was detected ventrally in the motor column (VH; Fig. 4B) and in a population of commissural neurons positioned dorsolaterally (Fig. 4B). srGAP2 was also detected in the region of the roof plate (RP; Fig. 4B) and floor plate (FP; Fig. 4B) and in the DRGs (Fig. 4B). At E11.5 srGAP3 was expressed dorsally (Fig. 4C; arrows) and around the RP (Fig. 4C). srGAP3 was also expressed in a population of commissural neurons positioned medially in the cord and around the FP at the site of commissural axon crossing (Fig. 4C). Weaker expression was also detected in the ventral motor columns (VH; Fig. 4C) and in the DRGs (Fig. 4C).

thumbnail image

Figure 4. srGAP1, srGAP2, and srGAP3 mRNA expression during spinal cord development from embryonic day (E) 11.5 until E13.5. A: At E11.5, srGAP1 mRNA expression is restricted to the ventral horns (VH). B: srGAP2 mRNA was detected in dorsolateral regions of the cord, including the VH and in the dorsal root ganglia (DRGs). srGAP2 mRNA was also detected in the region of the roof plate (RP) and the floor plate (FP) at this stage. C: srGAP3 mRNA was strongly detected in a stream of neurons running dorsoventrally as well as in the region of the FP. Expression was also more weakly detected in the DRGs, in a dorsal region of the cord around the RP, and in the VC. D: At E12.5, srGAP1 mRNA was located exclusively in the dorsal horns (DH) and again in the VH. E: At E12.5, srGAP2 mRNA was ubiquitously expressed throughout the entire cord and DRGs. Expression appeared slightly elevated in the medial part of the cord (asterisk). F: srGAP3 mRNA was strongly expressed throughout the cord and DRGs at E12.5; expression was particularly strong in the ependymal layer (EL). G: At E13.5, srGAP1 mRNA expression was still restricted to the DH and VH. H: srGAP2 mRNA was expressed ubiquitously throughout the cord and DRGs at E13.5. I: At E13.5, srGAP3 mRNA was expressed throughout the cord and DRGs and expression appeared to be stronger in the EL. All sections are sagittal. Scale bars = 100 μm.

Download figure to PowerPoint

DRG sensory axons enter the cord at E12.5, where they bifurcate and project branches toward ventral motor pools. Expression of srGAP1 mRNA remained in the ventral motor pools at E12.5 (VH; Fig. 4D) but extended to a subset of neurons in the dorsal horn (DH; Fig. 4D), where DRG sensory axons enter the cord. Both srGAP2 and srGAP3 mRNA localized ubiquitously throughout the cord and DRGs at E12.5 (Fig. 4E,F). srGAP3 mRNA appeared to be more concentrated in the ependymal layer at this stage (EL; Fig. 4F) and srGAP2 expression appeared slightly higher in a region of the dorsoventral boundary (Fig. 4E; asterisk). At E13.5, srGAP1 remained restricted to the VH and DH (Fig. 4G), while srGAP2 was ubiquitously expressed throughout the cord and DRGs (Fig. 4H). srGAP3 was also expressed throughout the cord and DRGs (Fig. 4I), but stronger expression was observed in the EL (Fig. 4I).

srGAP expression during development of the olfactory system

Slit-Robo signaling is involved in the development of olfactory tracts; thus, we investigated the expression of srGAP mRNAs in the olfactory system. At E11.5 both srGAP2 (Fig. 1B; Olf) and srGAP3 (Fig. 1C; Olf) were weakly expressed in the olfactory pit, while srGAP1 mRNA was not detected (Fig. 1A; Olf). Coronal sections of E14.5 heads revealed that sgrGAP1, srGAP2, and srGAP3 were all expressed in the mitral cell layer (ML) of the olfactory bulb (Fig. 5A–C). srGAP1 and srGAP3 were also weakly detected in some scattered neurons of the GL (Fig. 5A,C). The olfactory epithelium contains olfactory sensory neurons, which project axons to the olfactory bulb. srGAP2 mRNA was strongly expressed throughout the entire olfactory epithelium (Fig. 5B; arrows), whereas srGAP1 and srGAP3 mRNA was not detected in the olfactory epithelium (Fig. 5A,C). srGAP1, srGAP2, and srGAP3 were all expressed in the vomeronasal organ (VN; Fig. 5A–C) and srGAP2 appeared to be the most strongly expressed. At P1, srGAP1, 2, and 3 were all still expressed in the ML of the olfactory bulb (Fig. 5D–F) and srGAP1 and three were also detected in the GL of the olfactory bulb (Fig. 5D–F). srGAP2 was, again, strongly detected in the olfactory epithelium at P1 (Fig. 5E; arrows). srGAP1 and srGAP3 were still not detected in the olfactory epithelium at this stage (Fig. 5D,F).

thumbnail image

Figure 5. srGAP expression in the developing olfactory system at embryonic day (E) 14.5 and postnatal day (P) 1. A: srGAP1 mRNA was detected in the olfactory bulb at E14.5; expression appeared to be concentrated in the mitral cell layer (ML). srGAP1 mRNA was also weakly detected in the vomeronasal organ (VN). B: srGAP2 mRNA expression was observed in the ML of the olfactory bulb at E14.5. Expression was also strongly detected in the VN and throughout the olfactory epithelium (arrows) surrounding the nasal cavity (NC). C: srGAP3 mRNA was expressed in the VN and in the olfactory bulb, particularly the ML. D: srGAP1 was strongly expressed in the ML of the olfactory bulb at P1. Expression was also detected in the granule layer (GL). E: srGAP2 mRNA was detected in the ML at P1. Expression was also detected in the olfactory epithelium (arrows). F: At P1, srGAP3 was expressed in the ML and more weakly in the GL of the olfactory bulb. Sections A–C are coronal, D–F represent sagittal sections. Scale bars = 500 μm.

Download figure to PowerPoint

srGAP expression during eye development

Retinal ganglion axons require Slit-Robo signaling for guidance within the optic fiber tract (Thompson et al.,2006); therefore, we studied srGAP mRNA expression in the developing eye. We found that srGAP1 was not expressed in the neural layers of the eye at E11.5 (Fig. 1A; OP), E12.5 (Fig. 6A), or E13.5 (data not shown, see Table 1). srGAP1 was first very faintly detected in the retinal ganglion layer (RGL) at E14.5 (data not shown, see Table 1) and was also weakly expressed exclusively in the RGL at E16.5 (Fig. 6D). srGAP2 mRNA was detected in the optic pit at E11.5 (OP; Fig. 1B) and was strongly detected exclusively in the RGL at E12.5 (Fig. 6B), E13.5, E14.5 (data not shown), and E16.5 (Fig. 6E). srGAP3 mRNA was expressed in the OP at E11.5 (Fig. 1C) and from E12.5 until E16.5, expression was detected in both the RGL and the surrounding neuroblastic layer (NBL; Fig. 6C,F).

thumbnail image

Figure 6. srGAP1, srGAP2, and srGAP3 in the developing eye at embryonic day (E) 12.5 (A–C) and E16.5 (D–F). A: srGAP1 mRNA was not detected in the developing eye at E12.5, either in the retinal ganglion cell layer (RGL) or the neuroblastic layer (NBL). D: Weak srGAP1 expression was detected in the RGL at E16.5. B: At E12.5 srGAP2 mRNA was detected in the RGL of the developing eye, but was absent from the NBL. E: This pattern of expression was maintained at E16.5. C: srGAP3 mRNA was detected throughout the RGL and NBL of the developing eye at E12.5. F: This pattern of expression was maintained at E16.5. Scale bars = 100 μm. All sections are coronal.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED

Our knowledge of the function of the srGAP genes in neural development is in its infancy. Given that the srGAP genes regulate RhoGTPase function (Wong et al.,2001; Endris et al.,2002; Soderling et al.,2002) and srGAP3 binds to WAVE1, a regulator of actin polymerization (Soderling et al.,2002,2007), the principal role is likely to be in the structural reorganization of the neuronal cytoskeleton for directional migration and axon projection. To gain more insight into the functions of the srGAP family, we have extensively described the expression of the three known srGAP genes during the development of the murine nervous system. Strikingly, we found that srGAP genes are expressed in many CNS tissues throughout embryonic and early postnatal development. As Slit-Robo signaling is implicated in neurogenesis of multiple CNS tissues, our results are consistent with a role for srGAP genes in these signaling pathways.

Roles for srGAP genes in neurogenesis

Neurogenesis refers to the birth of neuronal precursors and occurs primarily in ventricular zones, although subventricular zones are sites of secondary neurogenesis at later developmental stages (Sanes et al.,2006). Slit-Robo signaling was recently implicated in the division of cortical interneuron precursors (Andrews et al.,2008), which is interesting, as we consistently observed srGAP3 expression in neuroepithelial layers of the developing brain, such as the hippocampal neuroepithelium, neuroepithelium of the superior and inferior colliculi, the neuroblastic layer of the retina, and the ventricular zone of the cerebral cortex. This consistency of expression at sites of proliferation points to a role for srGAP3 in Slit-Robo-mediated neurogenesis. Furthermore, we observed a decrease in srGAP3 expression in the ventricular neuroepithelium of the superior colliculus after the peak of neurogenesis at E13.5, which implies that srGAP3 is involved in the proliferation of precursors in the superior colliculus.

In contrast, srGAP2 was strikingly absent from sites of neurogenesis, but often strongly detected in regions of neuronal migration and differentiation, such as the retinal ganglion layer and the differentiation zones of the superior and inferior colliculi. We also observed that srGAP2 was distinctly absent from the dentate gyrus of the postnatal hippocampus, the site of hippocampal neurogenesis in the adult brain (Kempermann et al.,2004).

While srGAP3 was strongly detected in neuroepithelial layers and srGAP2 in early zones of differentiation, we observed that the onset of srGAP1 expression was often later, after primary neurogenesis. This late onset of srGAP1 expression compared to srGAP2 and srGAP3 was observed in the retinal ganglion layer, the cerebral cortex, the hippocampus, the granule layer and Purkinje layer of the cerebellum, parts of the olfactory system, and the superior and inferior colliculi (see Table 1). This implies that srGAP1 is dispensable for earlier stages of neuronal development and is perhaps more crucial in later stages, such as axon guidance or in postnatal neurogenesis and cell migrations. Indeed, srGAP1 has been implicated in the Slit-mediated repulsive migration of neurons from the subventricular zone (Wong et al.,2001) toward the olfactory bulb: the so-called rostral migratory stream, which persists into adulthood (Ghashghaei et al.,2007). Also, srGAP1 mRNA, while distinctly absent from the hippocampal neuroepithelium, was strongly expressed in the postnatal dentate gyrus.

Are srGAPs involved in neuronal migration?

Neuronal migration in the developing brain is both complex and diverse. Slit-Robo signaling has been implicated in many migratory pathways, including migration of precerebellar neurons (Gilthorpe et al.,2002), cortical interneuron migration (Andrews et al.,2006,2008), migration of olfactory interneuron precursors to the olfactory bulb (Hu,1999; Nguyen-Ba-Charvet et al.,2004), and the migration of cerebral cortical neurons from the subventricular zone to the cortical plate (Hu,1999). srGAPs may be involved in these signaling pathways, regulating RhoGTPase activity to mediate cytoskeletal reorganization for directional migration, as has been demonstrated for srGAP1 in subventricular zone cell migration (Wong et al.,2001).

Between E11.5 and E14.5 in the developing mouse brain, a subset of cortical interneuron precursors migrates from the medial ganglionic eminence toward the neocortex (Anderson et al.,2001). Cortical interneurons enter the cortex and migrate predominantly along the intermediate zone and ventricular zone in later stages of development. All three srGAP genes were detected in these layers at E16.5, consistent with a role for srGAP signaling in guided migration of later interneuron cohorts. It is also possible that the expression of srGAP genes in the cortex reflects a role in the radial migration of neuronal precursors from the ventricular zone to their target layers within the cortex. Colabeling with specific markers for distinct cortical neuron populations is required to clarify this.

Are srGAP genes involved in axon guidance and branching?

Slit-Robo signaling is required for normal commissural axon crossing in the murine spinal cord (Long et al.,2004; Sabatier et al.,2004; Chen et al.,2008). We found that srGAP2 and srGAP3 were expressed in distinct populations of commissural neurons at E11.5, when their axons were first crossing the midline. srGAP2 was expressed in commissural neurons with their cell bodies in more lateral positions, while srGAP3 mRNA was found in a population with their cell bodies in medial positions, closer to the floor plate. Interestingly, the expression pattern we observed for srGAP2 mRNA in the spinal cord was very similar to that of Robo2 (Long et al.,2004), while srGAP3 appeared to be expressed in the same population of commissural neurons as previously described for Robo1 (Long et al.,2004). Both Robo1 and Robo2 are thought to be involved in Slit-mediated repulsion from the midline after crossing in medially positioned and laterally positioned neurons, respectively (Long et al.,2004). Slit-Robo1-srGAP3 signaling may repel commissural axons that project longitudinally in the ventral funiculus, while Slit-Robo2-srGAP2 signaling may mediate the repulsion of commissural axons which project longitudinally in the lateral funiculus. Indeed, Robo1−/− mice have an enlarged lateral funiculus, as more axons are projected further from the floor plate, whereas Robo2−/− mice have an enlarged ventral funiculus, as fewer axons are repelled to the lateral funiculus (Long et al.,2004).

Another aspect of spinal cord development is the projection of DRG sensory axons into the cord and projection of axon branches to ventral motor pools. The N-terminal portion of Slit2 was identified as a positive regulator of sensory axon elongation and branching (Wang et al.,1999) and both Slit1/Slit2 and Robo1/Robo2 double mutants demonstrate misprojection of bifurcating DRG sensory axons into the dorsal portion of the spinal cord (Ma and Tessier-Lavigne,2007). At E11.5, prior to collateral branching of DRG sensory neurons into the spinal cord, we demonstrated that srGAP1 mRNA was expressed exclusively in the ventral horns of the spinal cord, an expression pattern very similar to that described for Slit2 by Wang et al. (1999). At E12.5, when the first DRG sensory axons begin to enter the cord, we found that expression of srGAP1 mRNA extended to the dorsal horns of the spinal cord, in regions where sensory axons enter the cord. This extension of expression from the ventral to dorsal horns as development progresses was also described for Slit2 (Wang et al.,1999). This expression pattern was also maintained at E13.5, a time when later DRG sensory axon projections were entering the cord. Our data imply that srGAP1 may be involved in Slit2-Robo2 signaling pathways regulating DRG sensory axon elongation and branching.

Olfactory receptor neurons in the olfactory epithelium project axons toward specific glomeruli in the olfactory bulb. Robo2 but not Robo1 is strongly expressed in the olfactory epithelium (Marillat et al.,2002; Nguyen-Ba-Charvet et al.,2008) and Slit-Robo signaling is required for correct targeting of olfactory sensory axons to specific regions in the olfactory bulb (Nguyen-Ba-Charvet et al.,2008). We found that srGAP2 was expressed in the olfactory epithelium, while srGAP1 and srGAP3 were not, implicating srGAP2 in Slit-Robo mediated guidance of olfactory sensory axons to the olfactory bulb.

CONCLUSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED

In our study we have demonstrated that srGAP genes, like Slit and Robo genes, have diverse patterns of expression that are often distinct from each other and are therefore likely to be important for many aspects of CNS development. We purposely eliminated mature developmental stages from our study, as the online Allen Brain Atlas, provided by the Allen Institute for Brain Science, provides an excellent and detailed analysis of srGAP1, srGAP2, and srGAP3 gene expression in the adult mouse brain (www.brain-map.org). We have discussed our results in relation to Slit-Robo signaling, although srGAP proteins are likely to have other interaction partners that influence their function. Indeed, we already know that srGAP3 binds WAVE1 to regulate synaptogenesis in hippocampal neurons (Soderling et al.,2002,2007). Another point of consideration is that srGAP proteins appear to preferentially regulate different Rho-GTPases. srGAP1 was shown to predominantly regulate Cdc42 (Wong et al., 2002), while srGAP3 principally regulates Rac1 signaling (Endris et al.,2002; Soderling et al.,2002). As Rho-GTPases are known to have diverse roles in neurogenesis (Newey et al.,2004), the specificity of Rho-GTPase regulation by srGAP proteins will certainly have functional consequences.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED

The authors thank Dr. Kerry Tucker for critical reading of the article. All imaging was carried out in the Nikon Imaging Centre at the University of Heidelberg.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSION
  7. Acknowledgements
  8. LITERATURE CITED