Crk and Crk-like play essential overlapping roles downstream of disabled-1 in the Reelin pathway

Abstract

Reelin controls neuronal positioning in the developing brain by binding to the two lipoprotein receptors, very-low-density lipoprotein receptor and apolipoprotein E receptor 2, to stimulate phosphorylation of Disabled-1 (Dab1) by the Fyn and Src tyrosine kinases. Crk and Crk-like (CrkL) have been proposed to interact with tyrosine phosphorylated Dab1 to mediate downstream events in the Reelin pathway. However, these adaptor proteins are widely expressed, and they fulfill essential functions during embryonic development. To address their specific roles in Reelin-mediated neuronal migration, we generated mutant mice, by Cre-loxP recombination, lacking Crk and CrkL in most neurons. These animals displayed the major anatomic features of reeler including, cerebellar hypofoliation, failure of Purkinje cell migration, absence of preplate splitting, impaired dendritic development, and disruption of layer formation in the hippocampus and cerebral cortex. However, proximal signaling involving tyrosine phosphorylation and turnover of Dab1 occurred normally in the mutant mouse brain and in primary cortical neurons treated with Reelin. In contrast, two downstream signaling events, Reelin-induced phosphorylation of C3G and Akt, were not observed in the absence of Crk and CrkL in mouse embryonic cortical neurons. These findings place C3G and Akt phosphorylation downstream of Crk and CrkL, which play essential overlapping functions in the Reelin signaling pathway.

Figure 1.

Targeted disruption of Crk and CrkL (dcKO) in mice. A, Schematic diagram of WT and homologous recombinant (floxed) alleles of CrkL. Filled triangles represent loxP sites. The relevant restriction sites XbaI (X), BstBI (BB), AatI (A), PacI (P), ClaI (C), and BsmI (B) are indicated. Colored bars indicate the locations of Southern probes (thick) and DNA fragments (thin) detected by the probes. B, Southern blot analysis of CrkL floxed mice. Genomic DNA from mouse tail was digested with XbaI and BsmI and hybridized with 5′- and 3′-external CrkL probes, respectively. C, Growth curves of dcKO mice and WT littermates. Body weight (in grams) of mice was measured two or three times a week, and data from females (27 dcKO and 35 WT are shown here) and males (data not shown) were analyzed separately, and revealed similar trends. Each point represents the mean ± SEM. D, Photograph of a 4-week-old dcKO mouse and a WT littermate. E, Image of perfused brains from a 4-week-old dcKO mouse and a wild-type littermate. Note the cerebellum is much smaller in dcKO than WT (see arrow). Scale bars: D, 1 cm; E, 0.5 cm.

Figure 2.

Histological analysis of WT and dcKO cerebellum from 4-week-old mice. A–D, Crk immunostaining of low (A, B) and high (C, D) resolution images of WT (A, C) and dcKO (B, D). E–H, CrkL immunostaining of low (E, F) and high (G, H) resolution images of WT (E, G) and dcKO (F, H). I–L, H&E staining of low (I, J) and high (K, L) resolution images of WT (I, K) and dcKO (J, L). Ectopic granule cells are indicated as yellow arrows. Red arrow indicates boundary between the igl and ml layers is not clear. M–P, Calbindin immunostaining of low (M, N) and high (O, P) resolution images of WT (M, O) and dcKO (N, P). Double-headed arrows indicate thickness of the molecular layers. Black and blue arrows indicate Purkinje cells that failed to migrate and those that partially migrated, respectively. The red arrow indicates Purkinje cells that reached their final destination. ml, Molecular layer; igl, internal granule cell layer; e-gl, ectopic granule cell layer. Scale bars: A, E, I, M, 500 μm; C, G, K, O, 50 μm.

Figure 3.

Histological analysis of WT and dcKO hippocampus from 4-week-old mice. A–D, Crk immunostaining of low (A, B) and high (C, D for CA1) resolution images of WT (A, C) and dcKO (B, D). E–H, CrkL immunostaining of low (E, F) and high (G, H for CA1) resolution images of WT (E, G) and dcKO (F, H). Arrow in F indicates scattered CrkL expression in the CA3 region of dcKO mice. I–N, H&E staining of low (I, J) and high (K, L for CA1; M, N for dentate gyrus) resolution images of WT (I, K, M) and dcKO (J, L, N). O, P, Ki67 immunostaining with hematoxylin (H) counterstaining of WT (O) and dcKO (P). Proliferating subgranule cells in the dentate gyrus are indicated as yellow arrows (M–P). Q–T, Map2 immunostaining of low (Q, R) and high (S, T for CA1) resolution images of WT (Q, S) and dcKO (R, T). spb, Suprapyramidal blade; ipb, infrapyramidal blade; h, hilus; so, stratum oriens; sp, stratum pyramidale; sr, stratum radiatum; gl, granule cell layer; sgl, subgranular cell layer. Scale bars: A, E, I, 500 μm; O, Q, 200 μm; C, G, K, S, 50 μm.

Figure 4.

Histological analysis of WT and dcKO cerebral cortex from 4-week-old mice. A–D, Immunostaining with anti-Crk (A, B) and anti-CrkL (C, D). E–H, H&E staining. High-resolution images of E (WT) and F (dcKO) are shown in G and H, respectively. The marginal zone (mz) is indicated by a yellow double-sided arrow. I–L, Map2 immunostaining. Neuronal cell bodies under the marginal zone or under the pial surface are indicated by black arrows in I and J. High-resolution images of I and J are shown in K and L, respectively, and apical dendrites of pyramidal neurons are marked depending on their orientation (radial, black arrows; random, blue arrows). M–P, Calbindin immunostaining. Layer-specific predominant staining is indicated by double-sided arrows and staining of dispersed interneurons is indicated by black arrows. High-resolution images of M and N are shown in O and P, respectively. mz, Marginal zone; spp, superplate. Scale bars: A, C, E, I, M, 250 μm; G, K, O, 50 μm.

Figure 5.

Reduced expression of Crk and CrkL in E15.5 dcKO cortex. Sagittal paraffin sections of embryonic cortex were immunostained with anti-Crk (A, B) and anti-CrkL (C, D) antibodies. Double-headed arrow in B indicates very weak expression of Crk in superficial layers. mz, Marginal zone; cp, cortical plate; sp, subplate; iz, intermediate zone; vz, ventricular zone. Scale bar, 100 μm.

Figure 6.

Histological analysis of E15.5 dcKO cortex. A–I, Every tenth sagittal paraffin section of whole embryos was stained with H&E (A–C), and representative images of the cortex were taken from the sections. Matching sections were processed for immunostaining. White arrows in B and C indicate aberrant cell-free rifts. Black arrow in D indicates subplate stained with anti-Map2 antibody, while black arrow in E indicates a heterotopia formed in the marginal zone of dcKO cortex. Black double-sided arrows in G–I indicate layers of strong calbindin immunoreactivity. J–L, Birthdating analysis. BrdU was administered to pregnant females at E11.5 to label subplate neurons. Locations of the cell populations were analyzed at E17.5. Black double-sided arrows in J–L indicate layers of strong BrdU immunoreactivity. mz, Marginal zone; cp, cortical plate; sp, subplate; iz, intermediate zone; vz, ventricular zone; spp, superplate. Scale bar, 100 μm.

Figure 7.

Reelin signaling pathway in E15.5 dcKO cortex and primary cortical neurons. A–C, Immunostaining of mouse E15.5 cortex of WT, dcKO, and reeler with anti-Dab1 antibody. Black vertical bars indicate layers of strong Dab1 immunoreactivity. Neuronal cell bodies located between the ventricular and intermediate zones are indicated as black arrows for comparison of their Dab1 immunoreactivity. Scale bar, 100 μm. D, Western blot analysis of E15.5 WT and dcKO cerebral cortical hemispheres with anti-Dab1, anti-Crk, and anti-CrkL antibodies. Anti-ShpII was used as a loading control. The ratio of Dab1 to ShpII was calculated and normalized to the ratio in WT cortical extract. E, F, Immunoprecipitation (IP) of extracts from cerebral cortical hemispheres from E15.5 dcKO (E) or reeler (F), and their WT embryos with anti-Dab1 antibody followed by blotting with anti-phosphotyrosine (pY-Dab1, 4G10) and anti-Dab1 antibodies. Both the ratio of Dab1 protein itself and the ratio of tyrosine phosphorylated Dab1 to total Dab1 protein were calculated and normalized to the ratio in WT cortical extract. G, H, Mouse embryonic cortical neurons were prepared from WT and dcKO embryos, cultured, and stimulated with either mock or purified Reelin for 15 min (G) to see stimulation of signaling molecules or for 3 h (H) to detect degradation of Dab1. Total protein extract was subjected to IP with anti-Dab1 and anti-C3G antibodies to detect tyrosine phosphorylation of Dab1 and C3G, respectively. Total protein extract was also subjected to blotting with antibodies against phospho-Akt (serine 473), Akt, Crk, CrkL, and phospholipase Cγ1 (PLCr1) as a loading control. The arrow in G indicates location of tyrosine phosphorylated C3G. I, Schematic diagram of the role of Crk and CrkL in the Reelin signaling pathway. We suggest the Reelin pathway consists of two subroutines, specific subroutine (Subroutine S) and general subroutine (Subroutine G). Molecules in Subroutine S include the pathway-specific molecules such as Reelin, Reelin receptors VLDLR and ApoER2, and Dab1, while Subroutine G consists of common molecules such as Crk, CrkL, Akt, and Rap1. We suggest that the two subroutines operate separately, but are coupled to each other by the interaction of tyrosine phosphorylated Dab1 (pY-Dab1) with Crk/CrkL. PI3K, Phosphatidylinositol 3-kinase; Rap1, repressor activator protein; rl, reeler.