Collapsin response mediator protein 1 (CRMP1) is one of the CRMP family members that mediates signal transduction of axon guidance molecules. Here, we show evidence that CRMP1 is involved in Reelin (Reln) signaling to regulate neuronal migration in the cerebral cortex. In crmp1−/− mice, radial migration of cortical neurons was retarded. This phenotype was not observed in the sema3A−/− and crmp1+/−;sema3A+/− cortices. However, CRMP1 was colocalized with disabled-1 (Dab1), an adaptor protein in Reln signaling. In the Relnrl/rl cortex, CRMP1 and Dab1 were expressed at a higher level, yet tyrosine phosphorylated at a lower level. Loss of crmp1 in a dab1 heterozygous background led to the disruption of hippocampal lamination, a Reeler-like phenotype. In addition to axon guidance, CRMP1 regulates neuronal migration by mediating Reln signaling.
The development of the neocortex has been of keen interest, because the cortex is uniquely mammalian and forms the basis for higher function. The neocortex consists of six layers of neurons that have distinct morphological and functional identities. The development of these neuronal layers involves the migration of neurons to their final laminar positions (Bielas et al., 2004).
Reelin (Reln) is a secreted protein that is involved in neuronal migration (Lambert de Rouvroit and Goffinet, 1998). During Reln binding to ApoER2/VLDLR receptors, the cytoplasmic adaptor disabled-1 (Dab1) becomes phosphorylated on tyrosine residues by Src-type tyrosine kinases (Hiesberger et al., 1999; Howell et al., 1999a; Bock and Herz, 2003). Perinatal lethality caused by widespread defects in neuronal migration during CNS development is evident in mice deficient for cyclin-dependent kinase 5 (Ohshima et al., 1996). Although how these molecules regulate neuronal migration remains obscure, several intracellular molecules, including Lis 1 (PAFAH1b1), doublecortin, Nudel, microtubule associated protein (MAP)-1B, MAP-2, and the small GTPase Rac1, have been implicated as mediators of these molecules and are involved in cytoskeletal organization (Gupta et al., 2002).
Collapsin response mediator protein (CRMP) was identified as a signaling molecule of Sema3A (Goshima et al., 1995). CRMPs are now known to be composed of five homologous cytosolic proteins; all of the family proteins are phosphorylated and are highly expressed in developing and adult nervous systems (Wang and Strittmatter, 1996; Fukada et al., 2000; Inatome et al., 2000; Yuasa-Kawada et al., 2003). CRMPs also mediate other signaling, such as NT3, and are involved in many aspects of neuronal cell development by regulating cytoskeletal organization (Goshima et al., 1995, 2002; Quach et al., 2004; Arimura and Kaibuchi, 2005; Uchida et al., 2005; Yoshimura et al., 2005). However, in vivo roles of CRMPs are mostly unknown.
To elucidate in vivo roles of CRMP1 in the development of the CNS, we have generated crmp1−/− mice (Charrier et al., 2006). We demonstrated that CRMP1 mediates Sema3A signaling in vivo, which is involved in spine maturation (N. Yamashita, A. Morita, Y. Uchida, F. Nakamura, H. Usui, M. Taniguchi, J. Honnorat, P. Kolattukudy, N. Thomasset, K. Takei, T. Takahashi, and Y. Goshima, unpublished observations). Intense crmp1 expression is observed in the cerebral cortex at the embryonic day 16.5 (E16.5), the period of neuronal migration (Bielas et al., 2004). In the present study, we investigated whether CRMP1 plays a role in neuronal migration in the cerebral cortex. We found that neuronal migration was retarded in the crmp1−/− cerebral cortex. In embryonic Relnrl/rl cortex, CRMP1 was tyrosine phosphorylated at a lower level and expressed at a higher level when compared with Relnrl/+. Furthermore, loss of crmp1 in a dab1 heterozygous background led to the disruption of hippocampal lamination, a Reeler-like phenotype. We propose that CRMP1 is involved in radial neuronal migration through Reln signaling.
Materials and Methods
5-Bromo-2′-deoxyuridine (BrdU) was purchased from GE Healthcare (Piscataway, NJ). Anti-Dab1 (B3; rabbit polyclonal) antibody was a kind gift from B. Howell (National Institute of Neurological Disorders and Stroke, Bethesda, MD). Monoclonal antibodies against CRMP1 (2C6G and 2E7G) were raised as described previously (N. Yamashita, A. Morita, Y. Uchida, F. Nakamura, H. Usui, M. Taniguchi, J. Honnorat, P. Kolattukudy, N. Thomasset, K. Takei, T. Takahashi, and Y. Goshima, unpublished observations). Other antibodies used were anti-BrdU (mouse monoclonal; MBL, Nagoya, Japan), anti-β-actin (mouse monoclonal; Sigma, St. Louis, MO), anti-Myc (9E10; mouse monoclonal; Sigma), anti-MAP-2 (2a + 2b) (mouse monoclonal; Sigma), anti-Nestin (mouse monoclonal; Chemicon, Temecula, CA), anti-p-Tyr (PY99; mouse monoclonal; Santa Cruz Biotechnology, Santa Cruz, CA), Cy3-labeled goat anti-Armenian hamster (Jackson ImmunoResearch, West Grove, PA), and Alexa 488-labeled goat anti-rabbit (Invitrogen, Cergy Pontoise, France).
reeler B6C3Fe mice were from The Jackson Laboratory (Bar Harbor, ME). yotari mice were spontaneous mutants at the dab1 allele (Yoneshima et al., 1997). sema3A and crmp1 mutant mice were generated as described previously (Taniguchi et al., 1997; Charrier et al., 2006). Genotypes of the offspring of all mutant mice were assessed using PCR, as described previously (D'Arcangelo et al., 1997; Kojima et al., 2000; Sasaki et al., 2002; Charrier et al., 2006). Mice were housed in the standard mouse facility and fed with an autoclaved diet and water. All procedures were performed according to the guidelines outlined by the institutional Animal Care and Use Committee of the Yokohama City University Graduate School of Medicine. Throughout the experimental procedures, all efforts were made to minimize the number of animals used and their suffering.
Immunoblot analysis and immunoprecipitation.
Brain samples were homogenized in immunoprecipitation (IP) buffer [20 mm Tris-HCl, pH 8.0, 150 mm NaCl, 1 mm EDTA, 10 mm NaF, 1 mm Na3VO4, 1% Nonidet P-40, 50 μm ρ-APMSF (ρ-amidinophenylmethanesulfonyl fluoride), and 10 μg/ml of aprotinin]. The lysates were centrifuged at 1200 rpm for 15 min at 4°C, and supernatants were normalized for total protein concentrations. The samples were then used for immunoblot analysis of anti-CRMP1 (2E7G; 1:2500), anti-β-actin (1:5000), anti-Dab1 (1:500), anti-MAP-2 (2a + 2b) (1:2000), and anti-Nestin (1:1000) antibodies. To detect the tyrosine phosphorylation of CRMP1, the samples were incubated with 1 μg of anti-CRMP1 (2E7G) antibody overnight at 4°C, followed by additional incubation with protein G–Sepharose (GE Healthcare) for 2 h at 4°C. After washing three times with IP buffer, the samples were used for immunoblot analysis of anti-CRMP1 (2E7G) and anti-p-Tyr (1:5000) antibodies.
In vivo phosphorylation assay.
HEK293T cells were seeded at 2.0 × 105 cells/well in a six-well plate. One day later, the cells were transfected with CRMP1-Myc with or without the wild-type, constitutive-active or dominant-negative form of Fyn expression vector using Fugene6 transfection reagent (Roche, Meylan, France). After 24 h, cells were lysed by IP buffer and then immunoprecipitated with 1 μg of anti-Myc antibody. The samples were used for immunoblot analysis of anti-Myc (1:2500) and anti-p-Tyr antibodies.
Cryostat brain sections (15 μm thick) were treated with 0.1% Triton X-100 in TBS with Tween 20 (TBST) for 10 min at room temperature. Immunostaining was performed according to standard protocols using anti-CRMP1 (2C6G; 1:1000) and anti-Dab1 (1:100) antibodies. Hydrolyzed, paraffin wax-embedded sections (6 μm thick) were used for immunohistochemistry of anti-MAP-2 (2a + 2b) (1:200) and anti-Nestin (1:100) antibodies. Slides were analyzed using a laser-scanning microscope (LSM510) with a water-immersed objective at 40× (C-Apochromat/1.2W corr) equipped with an Axioplan 2 imaging microscope (Carl Zeiss, Jena, Germany).
For birth-dating analysis, cells were labeled with BrdU (30 mg/kg) in the E14.5 or E16.5 neocortical wall. Pups were killed at E18.5, postnatal day 3 (P3) or P10, and brains were fixed in 4% paraformaldehyde for paraffin wax-embedded sectioning. Hydrolyzed, paraffin wax-embedded sections (6 μm thick) were heat-treated at 120°C for 20 min in 10 mm sodium citrate buffer, pH 6.0. Sections were then treated with 0.1% Triton X-100 in TBST for 10 min and with 1.5N HCl for 30 min at room temperature. Immunostaining was performed according to standard protocols using anti-BrdU antibody (1:1000). Slides were analyzed using Olympus (Tokyo, Japan) IX-71 microscopy using a 10× objective. For quantitative measurements, comparable sections were chosen at the somatosensory cortex, which was divided into 10 horizontal bins from the superficial to the deep, and labeled nuclei in each bin were counted (Teng et al., 2001).
Data are shown as mean ± SEM. The statistical significance of the results was analyzed using a Student's t test.
Neuronal migration in crmp1−/− cortex
To examine neuronal migration, we performed BrdU birth-dating analysis in the crmp1−/− cerebral cortex. In the brains of P10 injected at E14.5 (P10–BrdU–E14.5) crmp1+/− cortex, most labeled neurons were positioned in the deep layers of the cerebral cortex destined to form layers IV and V. In crmp1−/− mice, the majority of labeled cells were positioned at layers IV and V. The percentage of labeled cells positioned in those regions, however, was decreased, and the percentage of other regions was increased (Fig. 1A). A similar phenotype was also observed in the P10–BrdU–E16.5 crmp1−/− cortex. Most labeled neurons were positioned in the upper layers of the cortex, which were destined to form layers II and III in crmp1+/− cortex. However, the percentage of correctly positioned neurons was decreased, whereas the percentage of the cells in the other regions tended to be increased in crmp1−/− mice (Fig. 1B). We also investigated E18.5–BrdU–E14.5 and P3–BrdU–E16.5 crmp1+/− and crmp1−/− mouse cortices to observe the radial migration. In both brains, the retarded phenotype of neuronal migration was observed in the crmp1−/− cortex (data not shown). We next investigated neuronal differentiation and survival in the crmp1−/− cortex. At E16.5, apoptotic cells were rarely observed by terminal deoxynucleotidyl transferase-mediated biotinylated UTP nick end labeling staining. There was no difference in the proliferation of cortical progenitor cells between crmp1−/− and crmp1+/− cortices by BrdU proliferative analysis (data not shown). Furthermore, immunoreactivity and protein expression levels of Nestin and MAP-2 (2a + 2b), markers of dividing precursors and of matured neurons, respectively, in E16.5 crmp1−/− mice, were not different from those in crmp1+/− mice (Fig. 1C,D). These results indicate that CRMP1 plays a major role in neuronal cells at the postmitotic phase.
We next investigated whether a Sema3A-CRMP1 signaling cascade was also involved in neuronal migration in the cerebral cortex. Because most of the sema3A−/− pups died after birth, we performed only E18.5–BrdU–E14.5 analysis in the sema3A−/− cortex. We did not observe any differences between wt, sema3A+/−, and sema3A−/− in the E18.5–BrdU–E14.5 cortices (Fig. 2A). In addition, double heterozygous crmp1 and sema3A mice did not show any differences compared with crmp1+/− mice in the E18.5–BrdU–E14.5, P10–BrdU–E14.5, P3–BrdU–E16.5, and P10–BrdU–E16.5 cortices (Fig. 2B,C) (data not shown). These results suggest that CRMP1 mediates some signaling molecules other than Sema3A in cortical layer formation.
CRMP1 is involved in Reln signaling
In radial migration in the cerebral cortex, tyrosine phosphorylation of Dab1 by a Reln signaling cascade has been shown to be essential. Biochemical analysis has revealed a decreased tyrosine phosphorylation of Dab1 in the Relnrl/rl cortex (Howell et al., 1999a, 2000). One of the tyrosine kinases that is involved in phosphorylation of Dab1 is Fyn, an Src-type tyrosine kinase (Bock and Herz, 2003). Because we found that CRMP1 was tyrosine phosphorylated by Fyn in HEK293T cells (Fig. 3A), we assumed that CRMP1 was one of the substrates phosphorylated by tyrosine kinases through a Reln signaling cascade. To examine this hypothesis, we immunoprecipitated Relnrl/+ and Relnrl/rl brain lysates with anti-CRMP1 antibody and analyzed tyrosine phosphorylation levels of CRMP1 by anti-phospho-tyrosine antibody. The phosphorylation of CRMP1 at tyrosine residue(s) was detected in the brain lysate from Relnrl/+ mice. The level of tyrosine phosphorylation of CRMP1 was decreased in the Relnrl/rl cortex (Fig. 3B). The relative level of tyrosine phosphorylation of CRMP1 in Relnrl/rl normalized by the amount of immunoprecipitated CRMP1 in Relnrl/+ (100.0 ± 25.0) was 38.8 ± 5.2% (N = 3; p < 0.05). In addition, the expression of CRMP1 showed a higher level in Relnrl/rl than in Relnrl/+ mice. The CRMP1 expression was increased to a lesser extent than was the Dab1 expression in the brain lysates from Relnrl/rl mice (Fig. 3C). The relative amount of CRMP1 normalized by β-actin in Relnrl/+ and Relnrl/rl was 100.0 ± 11.8 and 155.8 ± 17.3%, respectively (N = 3; p < 0.05).
In Relnrl/rl cortex, reduced Dab1 expression is observed at superplate, whereas its increased expression is observed at cortical plate (Rice et al., 1998). We examined the protein expression pattern of CRMP1 in Relnrl/+ and Relnrl/rl cortices. Intense anti-CRMP1 and Dab1 immunoreactivities were observed at the marginal zone in Relnrl/+ cortex, whereas reduced immunoreactivities were observed at superplate in Relnrl/rl cortex. In addition, increased immunoreactivities of anti-CRMP1 and Dab1 at the cortical plate were also observed in Relnrl/rl cortex. Increased CRMP1 expression was observed in some neurons, whereas increased Dab1 expression was in almost all of the neurons of the cortical plate in Relnrl/rl mice (Fig. 3D).
We further attempted to analyze the genetic interaction with crmp1 and dab1; we used yotari mice, one of the spontaneous mutants of dab1. Nissl staining of the brain at P10 revealed a disrupted hippocampal lamination in the loss of crmp1 in a dab1 heterozygous background (crmp1−/−;dab1yot/+), one of the Reeler-like phenotypes (Fig. 3E). This phenotype was not observed in crmp1−/− and dab1yot/+ hippocampus (Fig. 3E) (Assadi et al., 2003), implying that CRMP1 and Dab1 were synergistically involved in Reln signaling for neuronal migration.
We provide, for the first time, evidence that CRMP1 is involved in radial neuronal migration through Reln signaling in the developing cerebral cortex. In Relnrl/rl cortex, CRMP1 as well as Dab1, an adaptor molecule in the transduction of the Reln signal, were expressed at a higher level, yet these were tyrosine phosphorylated at a lower level. Loss of crmp1 in a dab1 heterozygous background led to the disruption of hippocampal lamination, a typical Reeler-like phenotype.
In the crmp1−/− mouse cortex, retardation in neuronal migration was observed by BrdU birth-dating analysis when compared with the crmp1+/− cerebral cortex (Fig. 1) (data not shown). It is possible that Sema3A regulates neuronal migration in the developing cerebral cortex. Indeed, Sema3A regulates the tangential migration of GABAergic neurons from ganglionic eminence to the cerebral cortex (Marin et al., 2001; Tamamaki et al., 2003). Expression of nrp1 and sema3A were also observed in the cerebral cortex at E16.5 (Kawakami et al., 1996; Skaliora et al., 1998). However, it appears that Sema3A signaling is not involved in radial migration during the period of cortical layer formation, because the retarded phenotype in radial neuronal migration was not seen in either the sema3A−/− or double heterozygotes of crmp1 and sema3A mouse cortices (Fig. 2). Consistently, the overall expression pattern of nrp1 was different from that of crmp1 in the cerebral cortex at E16.5 (N. Yamashita, A. Morita, Y. Uchida, F. Nakamura, H. Usui, M. Taniguchi, J. Honnorat, P. Kolattukudy, N. Thomasset, K. Takei, T. Takahashi, and Y. Goshima, unpublished observations).
In contrast, the expression pattern of crmp1 was similar to that of dab1. We clearly showed that the expression of CRMP1 was increased in the Relnrl/rl cortex and reduced tyrosine phosphorylation of CRMP1 (Fig. 3B,C). Furthermore, the protein expression pattern of CRMP1 was also similar to that of Dab1 at the marginal zone and the cortical plate. The expression level of CRMP1 was decreased at superplate and increased at the cortical plate in the Relnrl/rl cortex, respectively (Fig. 3D). In adult Relnrl/rl cerebellum, the number of granule cells was reduced, and transient reduction of granule cells was observed in crmp1−/− cerebellum (Rice et al., 1998; Charrier et al., 2006). We found that loss of crmp1 in a dab1 heterozygous background led to the disruption of hippocampal lamination, a Reeler-like phenotype (Fig. 3E). Because the mutations in Reln pathway genes cause distinctive abnormalities in many laminated brain regions, known as the Reeler phenotype (D'Arcangelo et al., 1999; Hiesberger et al., 1999; Howell et al., 1999b), these findings strongly suggest that CRMP1 mediates Reln signaling to regulate neuronal migration.
We assumed that Reln regulates neuronal migration through tyrosine phosphorylation of CRMP1 because decreased tyrosine phosphorylation of Dab1 is observed in the Relnrl/rl cortex (Howell et al., 1999a, 2000), and animals expressing the nonphosphorylated Dab1 protein have a phenotype similar to the dab1 null mutant (Howell et al., 2000). Fyn and Fes tyrosine kinases have also been proposed to be signaling components of Sema3A (Mitsui et al., 2002; Sasaki et al., 2002). We found that when CRMP1 and Fyn were introduced into HEK293T cells, CRMP1 was phosphorylated at tyrosine residue(s) (Fig. 3A). We tried to detect phosphorylation of CRMP1 at tyrosine residue(s) in the cultured cortical neurons treated with Reln. Up to now, we have not yet obtained clear evidence for tyrosine phosphorylation of CRMP1 after Reln stimulation. The important finding is, however, that the level of tyrosine phosphorylation of CRMP1 was decreased in the Relnrl/rl cortex (Fig. 3B). Additional studies are required to delineate phosphorylation site(s) of CRMP1 and its functional significance in Reln signaling.
The presence of Reln correlates with reduced Dab1 protein levels during embryonic development (Rice et al., 1998; Howell et al., 1999a). Our study showed that the expression of CRMP1 was higher in Relnrl/rl cortex compared with Relnrl/+ mice (Fig. 3C), suggesting that Reln regulates CRMP1 protein levels in the cerebral cortex. This further indicates that CRMP1 possesses a biochemical property similar to Dab1. Interestingly, some neurons exhibited an increased immunoreactivity of CRMP1 in the Relnrl/rl cortical plate at E16.5, whereas Dab1 accumulation was observed in almost all neurons in the cortical plate (Fig. 3D). This result suggests that CRMP1 may regulate migration of a subset of the cortical neurons. The neurons exhibiting increased immunoreactivity of CRMP1 in the Relnrl/rl cortex were distributed throughout the cerebral cortex (Fig. 3D). Consistent with this observation, crmp1−/− showed a retarded migration phenotype in both the upper and deep layers of the cerebral cortex (Fig. 1A,B). These findings suggest that neurons in which CRMP1 may regulate their migration are composed of a heterogeneous subpopulation, being unrelated to a specific cortical layer. The characterization of the CRMP1-positive neurons remains to be determined. The expression pattern of another CRMP family member, crmp2, was almost similar to that of crmp1 (Wang and Strittmatter, 1996). Both CRMP1 and CRMP2 are involved in mediating Sema3A signaling (Uchida et al., 2005). These findings are consistent with a mild defect in neuronal migration in the crmp1−/− cortex, rather than in the Relnrl/rl cerebral cortex. Indeed, disruption of a single gene, having overlapping functions, sometimes does not result in any phenotypic defects in the mutant animals (Kuo et al., 2005). The combined disruption with other CRMP family genes will therefore tell us the exact roles of CRMP1 in the developing cerebral cortex.
This is the first evidence for an important role of CRMP1 in Reln signaling. It appears that the CRMP family is involved in cytoskeletal organization by receiving various extracellular cues that contribute to neuronal development and maturation.
This work was supported by Core Research for Evolutional Science and Technology of the Japan Science and Technology Corporation (K.T., Y.G.), Grants-in-Aid for Scientific Research in a Priority Area from the Ministry of Education, Science, Sports and Culture (Y.G.), and the Yokohama Medical Foundation (N.Y., Y.G.). We thank Dr. K. Nakajima and Dr. H. Tabata for helpful discussions about the analysis of neuronal migration. We thank Dr. M. Ogawa for helpful discussions about the analysis of Reelin signaling cascade. We thank Dr. B. Howell for his kind gift of antibody. We also thank Y. Sugiyama, T. Okada, T. Ueda, T. Hida, K. Ugajin, and M. Miyazaki for technical assistance.
- Correspondence should be addressed to Yoshio Goshima, Department of Molecular Pharmacology and Neurobiology, Yokohama City University Graduate School of Medicine, Fuku-ura 3-9, Kanazawa Ward, Yokohama City 236-0004, Japan.