Advertisement

Involvement of the Phosphatidylinositol 3-Kinase/Rac1 and Cdc42 Pathways in Radial Migration of Cortical Neurons*

  • Daijiro Konno
    Affiliations
    Department of Neuroscience (D13), Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan
    Search for articles by this author
  • Saori Yoshimura
    Affiliations
    Department of Neuroscience (D13), Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan
    Search for articles by this author
  • Kei Hori
    Affiliations
    Department of Neuroscience (D13), Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan
    Search for articles by this author
  • Hisato Maruoka
    Affiliations
    Department of Neuroscience (D13), Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan
    Search for articles by this author
  • Kenji Sobue
    Correspondence
    To whom correspondence should be addressed. Tel.: 81-6-6879-3680; Fax: 81-6-6879-3689;
    Affiliations
    Department of Neuroscience (D13), Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan
    Search for articles by this author
  • Author Footnotes
    * This work was supported by Grant-in-aid for Scientific Research 15GS0312 from the Ministry of Education, Science, Sports and Culture of Japan (to K. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Open AccessPublished:November 22, 2004DOI:https://doi.org/10.1074/jbc.M408251200
      During cortical development, newly generated neurons migrate radially toward their final positions. Although several candidate genes essential for this radial migration have been reported, the signaling pathways regulating it are largely unclear. Here we studied the role of phosphatidylinositol (PI) 3-kinase and its downstream signaling molecules in the radial migration of cortical neurons in vivo and in vitro. The expression of constitutively active and dominant-negative PI 3-kinases markedly inhibited radial migration. In the neocortical slice culture, a PI 3-kinase inhibitor suppressed the formation of GTP-bound Rac1 and Cdc42 and radial migration. Constitutively active and dominant-negative forms of Rac1 and Cdc42 but not Akt also significantly inhibited radial migration. In migrating neurons, wild-type Rac1 and Cdc42 showed different localizations; Rac1 localized to the plasma membrane and Cdc42 to the perinuclear region on the side of the leading processes. These results suggest that both the PI 3-kinase/Rac1 and Cdc42 pathways are involved in the radial migration of cortical neurons and that they have different roles.
      During mammalian cortical development, neurons born in the cortical ventricular zone migrate radially, resulting in the inside-out formation of the cortical plate, which is composed of six layers. Although the molecular bases of radial migration are largely unknown, extracellular ligands and intracellular cytoskeletal and signaling components have been shown to be involved. Reelin, a large extracellular molecule secreted from Cajal-Retzius cells, is involved in the radial migration of cortical neurons and formation of the cortical plate (
      • Rice D.S.
      • Curran T.
      ). Recent studies (
      • Gleeson J.G.
      • Walsh C.A.
      ,
      • Gupta A.
      • Tsai L.H.
      • Wynshaw-Boris A.
      ) analyzing human neurological disorders such as lissencephaly and double cortex have identified LIS1 and double cortin as key molecules required for patterned neuronal migration in the neocortex. LIS1 and doublecortin and their associated proteins, including Nude and Nudel, co-localize with microtubules and/or centrosome in various types of cells, including neurons, and control their dynamics, suggesting they play key roles in radial migration through nucleokinesis (
      • Morris N.R.
      • Efimov V.P.
      • Xiang X.
      ). Analyses of mutant mice have demonstrated roles for Cdk5 and its neuron-specific activator, p35, in radial migration (
      • Gleeson J.G.
      • Walsh C.A.
      ,
      • Gupta A.
      • Tsai L.H.
      • Wynshaw-Boris A.
      ). Moreover, Cdk5 phosphorylates doublecortin and/or focal adhesion kinase, altering their abilities to control microtubule dynamics (
      • Xie Z.
      • Sanada K.
      • Samuels B.A.
      • Shih H.
      • Tsai L.H.
      ,
      • Tanaka T.
      • Serneo F.F.
      • Tseng H.C.
      • Kulkarni A.B.
      • Tsai L.H.
      • Gleeson J.G.
      ). In addition, p35 binds directly to Rac1 in a GTP-dependent manner, and the Cdk5-p35 complex causes phosphorylation of Pak1, a downstream effector of Rac and Cdc42, in a Rac-dependent manner, resulting in the down-regulation of Pak1 kinase activity (
      • Nikolic M.
      • Chou M.M.
      • Lu W.
      • Mayer B.J.
      • Tsai L.H.
      ).
      Phosphatidylinositol (PI)
      The abbreviations used are: PI, phosphatidylinositol; PI(3,4,5)P3, phosphatidylinositol 3,4,5-triphosphate; GFP, green fluorescence protein; EGFP, enhanced green fluorescence protein; MAP2, microtubule associated protein 2; CP, cortical plate; VZ, ventricular zone; SVZ, subventricular zone; IZ, intermediate zone; GST, glutathione S-transferase, MTOC, microtubule organizing center.
      1The abbreviations used are: PI, phosphatidylinositol; PI(3,4,5)P3, phosphatidylinositol 3,4,5-triphosphate; GFP, green fluorescence protein; EGFP, enhanced green fluorescence protein; MAP2, microtubule associated protein 2; CP, cortical plate; VZ, ventricular zone; SVZ, subventricular zone; IZ, intermediate zone; GST, glutathione S-transferase, MTOC, microtubule organizing center.
      3-kinase, which produces phosphatidylinositol 3,4,5-triphosphate (PI[3,4,5]P3), is activated by several extracellular ligands, a process that is mediated through their receptor tyrosine kinases and G protein-coupled receptors (
      • Iijima M.
      • Huang Y.E.
      • Devreotes P.
      ,
      • Merlot S.
      • Firtel R.A.
      ). Several downstream signaling molecules of PI 3-kinase have been reported. In many kinds of cells, PI 3-kinase regulates cell polarization and migration via Rac, Cdc42, and/or Akt (protein kinase B). Rac1 and STEF/Tiam1, a Racspecific guanine nucleotide exchange factor, are involved in the radial migration of cortical neurons (
      • Kawauchi T.
      • Chihama K.
      • Nabeshima Y.
      • Hoshino M.
      ). However, the precise function of the PI 3-kinase/Rac1 and/or Cdc42 pathways in neuronal migration has not been reported.
      Here we demonstrate the involvement of the PI 3-kinase/Rac1 and Cdc42 pathways in the radial migration of cortical neurons in vivo and in vitro. The excess activation and the inactivation of PI 3-kinase, Rac1, and Cdc42, but not Akt (protein kinase B), markedly inhibited radial migration. Consistent with this, a PI 3-kinase inhibitor suppressed radial migration and reduced the formation of GTP-bound Rac1 and Cdc42 in neocortical slice culture. We further demonstrate the differential localization of Rac1 and Cdc42 in migrating neurons. These results suggest different roles for the PI 3-kinase/Rac1 and Cdc42 pathways in the radial migration of cortical neurons.

      EXPERIMENTAL PROCEDURES

      Plasmids—We used pCAGGS, a mammalian expression vector driven by chicken β-actin promoter, as described previously (
      • Niwa H.
      • Yamamura K.
      • Miyazaki J.
      ). To construct the expression vectors, the insert cDNAs were prepared from expression vectors for EGFP (pEGFPN1; Invitrogen), iSH2-p110α (gift from Dr. T. Katada), Δp85 (gift from Dr. I. Matsumura), Akt1-K179A (
      • Ohkawa Y.
      • Hayashi K.
      • Sobue K.
      ), Akt2-K181A, and Akt3-K177A (described below), and wild-type or mutant Rac1 and Cdc42 containing the Myc tag (gift from Dr. S. Narumiya) by PCR or enzymatic digestion. The Akt2 and Akt3 cDNAs were amplified by reverse transcription-PCR with human placenta mRNA and subcloned into the pCS vector. Point mutations changing lysine to alanine in Akt2 and Akt3 (Akt2-K181A and Akt3-K177A, respectively) were generated by site-directed mutagenesis. The insert cDNAs were blunted and cloned into pCAGGS. All constructs were confirmed by DNA sequencing.
      In Utero Electroporation—In utero electroporation was performed on ICR mice at embryonic day 15.5 (E15.5) as described previously (
      • Tabata H.
      • Nakajima K.
      ). ICR mice were provided by Japan SLC. In brief, expression vectors encoding GFP (0.2 μg/μl) and mutant proteins (described below) were co-injected into the lateral ventricle of the embryonic brain and co-electroporated with an Electro Square Porater BTX 830 (BTX). In all experiments except those using Δp85 and mutant forms of Akt, the expression vectors were used at a concentration of 0.2 μg/μl. The expression vectors encoding Δp85, Akt1-K179A, Akt2-K181, and Akt3-K177A were used at a concentration of 1.0 μg/μl. The brains were analyzed at 3 or 6 days after electroporation (E18.5 and P2, respectively).
      Immunohistochemistry—The dissected mouse brains were fixed in 0.1 m phosphate buffer (PB) containing 4% paraformaldehyde and 5% sucrose for 4 h at 4 °C. After fixation, the brains were cryoprotected with 25% sucrose in 0.1 m PB and embedded in OCT compound. Frontal sections (10-μm thick) were prepared from the anterior half of the cerebral cortex. The following procedures were carried out at room temperature. The sections were pretreated with blocking solution A (5% normal goat serum, 0.2% bovine serum albumin, 0.1% Triton X-100, 5 mm NaN3, 0.1 m phosphate-buffered saline) or B (1% bovine serum albumin, 0.1% Triton X-100, 5 mm NaN3, 0.1 m phosphate-buffered saline) for 1 h and incubated with primary antibodies for 16 h. A mouse monoclonal anti-MAP2 antibody (1:2000, clone HM-2, Sigma) was preincubated with the Alexa Fluor 546 secondary antibody (1:500, Molecular Probes) in blocking solution A. Rabbit polyclonal anti-Myc (1:200) and anti-p85 (1:200) and goat polyclonal anti-p110α (1:100) and anti-Akt (1:200) antibodies (Santa Cruz Biotechnology) were used as primary antibodies. These antibodies from rabbit and goat were diluted with blocking solutions A and B, respectively. All primary antibodies were detected with AlexaFluor 546 secondary antibodies. Fluorescence images were obtained using a Pascal 5 confocal laser-scanning microscope (Carl Zeiss) and then processed with Adobe Photoshop software. We acquired stacked images from 30 optical slices at 0.36-μm intervals, as shown in Fig. 4.
      Figure thumbnail gr1
      Fig. 1Constitutively active and dominant-negative PI 3-kinase and LY294002 inhibit the radial migration of cortical neurons. A, radial migration of cortical neurons was visualized by transfecting their progenitors with GFP using the in utero electroporation method. Electroporation was performed at E15.5 and analyzed at E18.5 (upper right three panels) or P2 (lower panels). Thin sections of the cerebral cortex were immunostained with anti-MAP2 antibody, a marker for post-mitotic neurons. An image showing the distributions of GFP fluorescence throughout the brain at P2 is shown in the upper left panel. Many GFP-positive cells migrated radially toward the CP at E18.5, and almost all the cells reached the upper region of the CP at P2. B and C, expression vectors encoding constitutively active (pCAGGS-iSH2-p110α) or dominant-negative (pCAGGS-Δp85) PI 3-kinase and GFP (pCAGGS-EGFP) were co-transfected into the VZ of embryonic brains using electroporation at E15.5, and coronal sections were analyzed at P2. Higher magnification views of the boxed region are shown with MAP2 immunofluorescence (upper right three panels). Immunostaining against p110α or p85 was performed to confirm the co-expression of GFP and iSH2-p110α (lower right three panels). D, quantitative analysis of A–C. The cerebral cortex was subdivided into five regions, indicated as CP1, CP2, CP3, IZ, and VZ/SVZ, and the number of GFP-positive cells and the frequency of cell position were counted. CP1, CP2, and CP3 indicate the upper, middle, and lower regions of the CP divided into three equal parts, respectively. E, the effect of LY294002, a specific inhibitor for PI 3-kinase, on the radial migration in neocortical slice culture. Prior to preparing the neocortical slices, in utero electroporation was performed at E14.5 to label the neurons with GFP. Neocortical slices were prepared at E16.0 and treated with 0.5% Me2SO or 25 or 200 μm LY294002. Images were taken at 2, 24, and 48 h and shown as GFP fluorescence images (green). After 48 h, nonviable cells were stained with propidium iodide (PI, red). Higher magnifications of the boxedregions in “48 h” were shown at its right side (control, 25 μm LY294002 and 200 μm LY294002) with merged images of GFP fluorescence (green) and propidium iodide staining (red). Compared with cells treated with control and 25 μm LY294002, almost all of cells treated with 200 μm LY294002 were nonviable as detected by propidium iodide staining. Arrows indicate the direction of GFP-positive migrating neurons. Bars, 50 μm (A–C) and 100 μm (E).
      Figure thumbnail gr4
      Fig. 4Differential distribution of Rac1 and Cdc42 in migrating neurons. The subcellular distribution of Rac1 and Cdc42 in migrating neurons. Expression vectors encoding Myc-tagged versions of wild-type Rac1 (pCAGGS-WT-Rac1) or Cdc42 (pCAGGS-WT-Cdc42) and GFP (pCAGGS-EGFP) were co-transfected into the neocortical VZ using in utero electroporation at E15.5, and coronal sections were analyzed at E18.5. Immunostaining against the Myc tag was performed to analyze the localization of Rac1 and Cdc42 in migrating neurons. Stacked images were constructed from 30 optical slices at 0.36-μm intervals (1st and 3rd panels), and higher magnification images located at the middle part of the stacked images (the boxed regions) are shown in the 2nd and 4th panels.
      Quantitative Analysis of Neuronal Migration—We subdivided the cerebral cortex into 5 regions (CP1, CP2, CP3, IZ, and VZ), counted the number of GFP-positive cells, and calculated the ratios of the GFP-positive cells to the total GFP-positive cells in the indicated regions as percentages with Excel software. In Figs. 1, 2, 3, CP1, CP2, and CP3 indicate the upper, middle, and lower regions of the cortical plate divided into three equal parts, respectively. Each value represents the average ± S.D. of the data from five independent embryos.
      Figure thumbnail gr2
      Fig. 2PI 3-kinase acts upstream of Akt, Rac1, and Cdc42 in the neocortex. A, the effect of LY294002 on the activation states of Akt, Rac1, and Cdc42. Neocortical slices were treated with 25 or 50 μm LY294002 for 16 h followed by GST-PAK pull-down assay. The amounts of Rac1, Cdc42, Akt, and phospho-Akt were determined by immunoblotting with anti-Rac1, -Cdc42, -Akt, and -phospho-Akt antibodies (left). The immunoblotting data were quantified from three independent experiments (right). B, expression vectors encoding dominant-negative Akt (pCAGGS-Akt1-K179A) and GFP (pCAGGS-EGFP) were co-transfected into the neocortical VZ using in utero electroporation at E15.5, and coronal sections were analyzed at P2. Higher magnification views of the boxed region are shown with MAP2 immunofluorescence (upper right three panels). Immunostaining against Akt was performed to confirm the co-expression of GFP and Akt1-K179A (lower right three panels). C, quantitative analysis of B. The number of GFP-positive cells and the frequency of cell position were counted described in . Bar, 50 μm.
      Figure thumbnail gr3
      Fig. 3Inhibition of radial migration by constitutive-active and dominant-negative Rac1 and Cdc42. A–D, expression vectors encoding constitutively active Rac1 (pCAGGS-V12Rac1) or Cdc42 (pCAGGS-V12Cdc42), dominant-negative Rac1 (pCAGGS-N17Rac1) or Cdc42 (pCAGGS-N17Cdc42), and GFP (pCAGGS-EGFP) were co-transfected into the neocortical VZ using in utero electroporation at E15.5, and coronal sections were analyzed at P2. The higher magnification views of the boxed region are shown with MAP2 immunofluorescence (upper right three panels). Immunostaining against the Myc tag was performed to confirm the co-expression of GFP and mutated Rac1 or Cdc42 (lower right three panels). E, quantitative analysis of A–D. The number of GFP-positive cells and the frequency of cell position were counted described in . Bar, 50 μm.
      Slice Culture of Neocortex—Slice culture of the neocortex was performed as described previously (
      • Hatanaka Y.
      • Murakami F.
      ) with minor modifications. In brief, VZ cells were transfected with a GFP expression vector by in utero electroporation at E14.5, and 300-μm-thick coronal slices were prepared from the dorsolateral part of the anterior half of the cerebral cortex at E16.0. Slices were placed on tissue culture membranes (Transwell-COL; Corning Glass) and cultured in growth medium (DMEM/F-12 containing N2 supplement and 10% fetal calf serum). After 2 h, 25, 50, or 200 μm LY294002, a specific inhibitor of PI 3-kinase, or 0.5% Me2SO was added to the medium. To investigate the resumption of neuronal migration, slices were replaced in the new medium without LY294002 at 48 h and then observed at 72 h. GFP fluorescence and bright phase images were obtained using a Pascal 5 confocal laser-scanning microscope (Carl Zeiss) at 2, 24, 48, and 72 h of culture. We analyzed the effect of LY294002 on the radial migration of GFP-positive cells from the IZ. To determine the viability of cells, the nonviable cells were labeled with propidium iodide (0.5 μg/ml) after 48 h with or without LY294002 treatment.
      GST-PAK Pull-down Assay—The Rac1 and Cdc42 activities were measured by pull-down assay by using GST-PAK-CRIB, which is a fusion protein of GST and the Cdc42/Rac-binding domain of PAK (
      • Paik J.H.
      • Chae S.
      • Lee M.J.
      • Thangada S.
      • Hla T.
      ). This method is used to measure the amount of GTP-bound Rac1 and Cdc42, because PAK only binds the GTP form of Rac1 and Cdc42. The prokaryotic GST-PAK-CRIB expression vector was a kind gift from Dr. S. Narumiya. The dorsal part of the cerebral cortex was taken from E16.0 ICR mouse brains, and cortical slices were prepared. These slices were cultured for 16 h with mild agitation in culture medium (Neurobasal medium supplemented with B27 and 0.5 mm glutamine) containing LY294002 (25 μm or 50 μm) or Me2SO (0.5%). Cultured slices were extracted with lysis buffer (10 mm MgCl2, 500 mm NaCl, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 50 mm Tris-HCl, pH 7.2). The extracts were incubated for 60 min at 4 °C with purified GST-PAK-CRIB produced by Escherichia coli that was immobilized on glutathione-Sepharose 4B (Amersham Biosciences). After the GST-PAK-CRIB-glutathione-Sepharose 4B was washed three times, the amounts of Rac1, Cdc42, and Akt1 were analyzed by immunoblotting using mouse monoclonal anti-Rac1 (Transduction Laboratories), rabbit polyclonal anti-Cdc42 (Santa Cruz Biotechnology), anti-phospho-Akt (Cell Signaling), and goat polyclonal anti-Akt (Santa Cruz Biotechnology) antibodies. We measured the intensities of corresponding bands by NIH Image J and further processed the data by using Excel and Cricket Graf software. Each value of quantitative data represents the average ± S.D. from three independent experiments.

      RESULTS

      To investigate the role of PI 3-kinase and its downstream signaling molecules in the radial migration of cortical neurons, we performed in utero electroporation to introduce expression vectors encoding GFP with or without signaling molecules into neuronal progenitor cells at the cortical VZ of embryonic brains at E15.5, and we then analyzed the radial migration of the labeled neurons in the cerebral cortex. When GFP expression vector (0.2 μg/μl) alone was introduced, many GFP-positive cells migrated to both the IZ and the lower region of the CP at E18.5 (Fig. 1A, upper panels). More than 80% of the total GFP-positive cells had reached the upper region of the CP, corresponding to layers II/III, on postnatal day 2 (P2) (Fig. 1A, lower panels). Fewer than 10% of the labeled cells remained within the IZ and the lower regions of the CP (Fig. 1, A and D). These migration patterns were not altered when the amount of GFP expression vector was varied (data not shown). Thus, exogenous GFP was useful for monitoring neuronal migration and did not itself interfere with the migration. Quantitative analysis are summarized in Table I.
      Table ISummary of quantitative analysis of radial migration of cortical neurons Each value represents the average pecentage ± S.D. of the data from five independent embryos.
      CP1CP2CP3IZVZ
      GFP83.8 ± 7.73.2 ± 2.82.7 ± 1.62.5 ± 2.57.9 ± 1.4
      iSH2-p110α2.1 ± 1.3
      p < 0.001
      14.1 ± 1.41
      p < 0.001
      28.0 ± 2.7
      p < 0.001
      48.0 ± 0.5
      p < 0.001
      6.7 ± 2.6
      Δp8523.3 ± 5.5
      p < 0.001
      20.5 ± 2.2
      p < 0.01 compared with GFP alone (Student's t test)
      15.5 ± 2.4
      p < 0.01 compared with GFP alone (Student's t test)
      25.0 ± 3.9
      p < 0.001
      15.5 ± 2.9
      p < 0.01 compared with GFP alone (Student's t test)
      Akt-K179A70.6 ± 4.75.5 ± 0.84.4 ± 0.85.6 ± 0.614.1 ± 4.1
      V12Rac116.1 ± 5.2
      p < 0.001
      10.7 ± 4.6
      p < 0.01 compared with GFP alone (Student's t test)
      10.8 ± 4.8
      p < 0.01 compared with GFP alone (Student's t test)
      40.87 ± 4.1
      p < 0.001
      24.1 ± 7.7
      p < 0.01 compared with GFP alone (Student's t test)
      N17Rac123.9 ± 5.5
      p < 0.001
      17.0 ± 6.1
      p < 0.01 compared with GFP alone (Student's t test)
      16.9 ± 3.9
      p < 0.01 compared with GFP alone (Student's t test)
      33.2 ± 5.6
      p < 0.001
      9.1 ± 3.4
      V12Cdc4257.4 ± 6.7
      p < 0.01 compared with GFP alone (Student's t test)
      10.7 ± 4.04.8 ± 1.712.6 ± 6.6
      p < 0.01 compared with GFP alone (Student's t test)
      13.8 ± 5.1
      N17Cdc4237.8 ± 4.7
      p < 0.001
      14.5 ± 4.0
      p < 0.01 compared with GFP alone (Student's t test)
      20.6 ± 0.4
      p < 0.001
      18.8 ± 3.9
      p < 0.001
      7.6 ± 0.6
      a p < 0.001
      b p < 0.01 compared with GFP alone (Student's t test)
      We next analyzed the expression of the catalytic subunits of class I PI 3-kinase during cortical development. Immunoblotting showed that the p110α and β subunits were intensely expressed throughout cortical development (data not shown). Thus, the major isoforms of PI 3-kinase expressed during cortical development are of class IA. We then examined the effect of class IA PI 3-kinases on the radial migration of cortical neurons by using constitutively active and dominant-negative forms of them in vivo (Fig. 1, B and C). Double immunolabeling for GFP and the mutant PI 3-kinases showed that GFP and the PI 3-kinase variants were co-expressed in almost all the transfected cells examined (Fig. 1, B and C). Therefore, GFP was used to identify the neurons expressing mutated PI 3-kinase. When a constitutively active form of PI 3-kinase, iSH2-p110α (iSH2-PI3K), and GFP were co-expressed, the radial migration was markedly suppressed. Fifty percent of the co-expressed cells were found in the IZ (Fig. 1, B and D). About 10% of the co-expressed cells remained within the VZ, which was similar to when the cells expressed GFP alone. The expression of a dominant-negative form of class IA PI 3-kinase, Δp85, also suppressed the radial migration. The labeled cells were diffusely distributed from the VZ/SVZ and the IZ to the CP (Fig. 1, C and D). By using neocortical slice culture, we further examined the effect of a PI 3-kinase inhibitor on the radial migration. In control slices, most of the GFP-positive cells remained within the IZ after 2 h in culture and had migrated to the CP after 24 h. Many GFP-positive cells migrated further, into the CP toward the pial surface, after 48 h (Fig. 1E, upper panels). When treated with a PI 3-kinase inhibitor (25 μm LY294002), the radial migration of the GFP-positive cells was slightly retarded after 24 h and further inhibited after 48 h (Fig. 1E, lower panels). A similar effect was observed using 50 μm LY294002 (data not shown). At the concentrations of LY294002 used here (25 and 50 μm), almost all of the GFP-positive cells were propidium iodide-negative, indicating that they are viable. Treatment with excess amounts of LY294002 (200 μm) markedly increased in PI-positive nonviable cells, in association with a dramatic decrease in the GFP-positive cells (Fig. 1E). Thus, LY294002 (25–50 μm) inhibited neuronal migration but did not affect the cell survival. Taken together, these results suggest that PI 3-kinase is required for the radial migration of cortical neurons.
      It has been well documented that PI 3-kinase regulates cellular events, including cell migration, through its downstream signaling molecules, which include Akt, Rac1, and Cdc42. Because immunoblotting showed the expression of Akt, Rac1, and Cdc42 throughout cortical development (data not shown), we investigated their roles in the radial migration of cortical neurons. We performed GST-PAK pull-down assays to examine the effect of LY294002 on the Akt, Rac1, and Cdc42 activities. When neocortical slices were treated with LY294002 (25 or 50 μm), the amounts of GTP-bound Rac1 and Cdc42 and phosphorylated Akt decreased in a LY294002 dose-dependent manner (Fig. 2A). These results suggested that Akt, Rac1, and Cdc42 act as possible downstream effectors of PI 3-kinase signaling during cortical development.
      Of these three molecules, we first examined the involvement of Akt on the radial migration. When dominant-negative Akt1 (Akt1-K179A) and GFP were co-expressed, the radial migration of GFP-positive cells was not affected. More than 70% of the cells migrated normally (Fig. 2, B and C). Similarly, dominant-negative Akt2 (Akt2-K181A) and Akt3 (Akt3-K177A) had no effect on the radial migration (data not shown). These results indicated that a major downstream target of PI 3-kinase, Akt, is not involved in the radial migration of cortical neurons.
      We next analyzed the roles of Rac1 and Cdc42 in the radial migration. When constitutively active Rac1 (V12Rac1) and GFP were co-expressed, the radial migration of GFP-positive cells was significantly suppressed, similar to the suppression seen in cells co-expressing the mutant PI 3-kinases. The GFP-positive cells co-expressing V12Rac1 showed a loss of leading processes. The radial migration of cells co-expressing dominant-negative Rac1 (N17Rac1) and GFP was also markedly retarded (Fig. 3, A and E). More than 40% of the GFP-positive cells remained within the IZ (Fig. 3, B and E). When constitutively active Cdc42 (V12Cdc42) and GFP were co-expressed, more than 60% of the GFP-positive cells migrated normally (Fig. 3, C and E). Only 12.6 ± 2.7% (mean ± S.E., n = 5) of the transfected cells remained in the VZ/SVZ and in the IZ, where they formed cell clusters. When dominant-negative Cdc42 (N17Cdc42) and GFP were co-expressed, the radial migration was inhibited; this effect was similar to, although not as strong as, that seen with N17Rac1 (Fig. 3, D and E). These results suggest that Rac1 and Cdc42 are downstream molecules of PI 3-kinase with different roles in the radial migration of cortical neurons.
      To investigate the different roles of Rac1 and Cdc42 in the radial migration, we compared the localization of wild-type Rac1 (WT-Rac1) and Cdc42 (WT-Cdc42) in radially migrating neurons in vivo (Fig. 4). The WT-Rac1 was specifically targeted to the plasma membrane of migrating neurons (Fig. 4, left panels). WT-Cdc42 was mainly localized to the perinuclear region on the side of the leading processes (Fig. 4, right panels). These results suggest that Rac1 and Cdc42 are differentially involved in the radial migration of cortical neurons, with roles in the regulation of leading process formation and cell polarity, respectively.

      DISCUSSION

      Postmitotic neurons generated in the VZ migrate into and form the CP via the following steps: release from the VZ, migration guided by radial glia, and positioning in the appropriate layers (
      • Gleeson J.G.
      • Walsh C.A.
      ,
      • Gupta A.
      • Tsai L.H.
      • Wynshaw-Boris A.
      ). As demonstrated here, mutant PI 3-kinase, Rac1, and Cdc42, but not Akt, inhibited the radial migration of cortical neurons. The PI 3-kinase mutants and/or GFP-transfected cells remaining within the VZ were only a minor population, suggesting that the PI 3-kinase pathways link to the radial glia-guided migration but not to release from the VZ. What might act upstream of PI 3-kinase? It has been well documented that integrin signaling through PI 3-kinase activation is involved in cell migration, invasion, and spreading (
      • Mercurio A.M.
      • Rabinovitz I.
      • Shaw L.M.
      ). The integrin family of cell-surface receptors is reported to function in the radial glia-guided neuronal migration through neuron-glia interactions (
      • Fishman R.B.
      • Hatten M.E.
      ,
      • Anton E.S.
      • Kreidberg J.A.
      • Rakic P.
      ). Laminin is transiently expressed on the cell surface of radial glial fibers during central nervous system development (
      • Liesi P.
      ), and inhibition of β1 integrins with their blocking antibodies induces the detachment of migrating neurons from radial glial fibers (
      • Fishman R.B.
      • Hatten M.E.
      ,
      • Anton E.S.
      • Kreidberg J.A.
      • Rakic P.
      ). Thus, neuronal integrins interacting with their ligands, such as laminin distributed along radial glial fibers, may act as an upstream activator of PI 3-kinase in the radial migration.
      PI 3-kinase is reported to have a role in chemotaxis. For example, PI 3-kinase and its product PI(3,4,5)P3 are involved in the chemotaxis of Dictyostelium and neutrophils (
      • Wang F.
      • Herzmark P.
      • Weiner O.D.
      • Srinivasan S.
      • Servant G.
      • Bourne H.R.
      ,
      • Weiner O.D.
      • Neilsen P.O.
      • Prestwich G.D.
      • Kirschner M.W.
      • Cantley L.C.
      • Bourne H.R.
      ,
      • Iijima M.
      • Devreotes P.
      ,
      • Funamoto S.
      • Meili R.
      • Lee S.
      • Parry L.
      • Firtel R.A.
      ). In these cells, PI(3,4,5)P3 is produced by the stimulation of chemoattractants and accumulates at the leading edge of the cells. Dictyostelium lacking PTEN, a PI 3-phosphatase, displays a broad irregular leading edge, with slow movement. These observations suggest that the local accumulation of PI(3,4,5)P3 at the proper regions is critical for chemotaxis in these cells. We demonstrated that both excess activation and inactivation of PI 3-kinase markedly suppressed the radial migration of cortical neurons in vivo. When cortical slices were treated with the PI 3-kinase inhibitor (LY294002, 25–50 μm), neuronal migration was also suppressed without cytotoxicity. We further investigated the reversibility of neuronal migration inhibited by LY294002. Neuronal migration was, however, irreversible (data not shown). This may be due to the possibility that once migrating neurons deviate from radial fibers, the cells might never migrate again. Our results taken together with the previous reports suggest that the proper activation of PI 3-kinase in regions that correspond to the critical regions in Dictyostelium and neutrophils may be play a similar role in the radial migration of cortical neurons as it does in the movement of non-neuronal cells.
      We demonstrated that Rac1 and Cdc42 act as downstream signaling molecules of PI 3-kinase in neuronal migration. The PI 3-kinase/Rac1 and/or Cdc42 pathways play important roles in cellular processes, including the maintenance of cell polarity and migration in mammalian cells (
      • Liu H.
      • Radisky D.C.
      • Wang F.
      • Bissell M.J.
      ,
      • Srinivasan S.
      • Wang F.
      • Glavas S.
      • Ott A.
      • Hofmann F.
      • Aktories K.
      • Kalman D.
      • Bourne H.R.
      ,
      • Keely P.J.
      • Westwick J.K.
      • Whitehead I.P.
      • Der C.J.
      • Parise L.V.
      ). The lipid product of PI 3-kinase, PI(3,4,5)P3, can greatly enhance the guanine nucleotide exchange factor activity for the Rho family GTPases, resulting in the activation of Rac and/or Cdc42 (
      • Welch H.C.
      • Coadwell W.J.
      • Stephens L.R.
      • Hawkins P.T.
      ). In addition, GTP-bound Rac1 and Cdc42 can directly bind p85, a regulatory subunit of PI 3-kinase, and activate PI 3-kinase (
      • Tolias K.F.
      • Cantley L.C.
      • Carpenter C.L.
      ,
      • Zheng Y.
      • Bagrodia S.
      • Cerione R.A.
      ), which in turn further increases the level of PI(3,4,5) P3. Thus, these direct and/or indirect activation pathways may provide a link between the PI 3-kinase and Rac1 and/or Cdc42 signaling pathways in the radial migration of cortical neurons.
      What does the differential distribution of Rac1 and Cdc42 in migrating neurons reflect? A previous study (
      • del Pozo M.A.
      • Price L.S.
      • Alderson N.B.
      • Ren X.D.
      • Schwartz M.A.
      ) reported the importance of the plasma membrane localization of Rac1 for cell adhesion in mammalian cells. Translocation of Rac to the plasma membrane is required for the activation of its effectors and cytoskeletal modification, including integrin-mediated cell adhesion to the extracellular matrix (
      • del Pozo M.A.
      • Price L.S.
      • Alderson N.B.
      • Ren X.D.
      • Schwartz M.A.
      ). Because we observed that neurons expressing mutant Rac1 lost their leading processes, the membrane localization of Rac1 may be required for its activation and for the formation of leading processes and/or their adhesion to radial glial fibers. The perinuclear localization of Cdc42 in migrating neurons probably corresponds to the centrosome (also called the microtubule organizing center (MTOC)) because of its positioning in the direction of migration. Cdc42 is also required for lysophosphatidic acid-induced MTOC orientation (
      • Palazzo A.F.
      • Joseph H.L.
      • Chen Y.J.
      • Dujardin D.L.
      • Alberts A.S.
      • Pfister K.K.
      • Vallee R.B.
      • Gundersen G.G.
      ) and the MTOC reorientation of wounded astrocytes (
      • Etienne-Manneville S.
      • Hall A.
      ). In humans, spontaneous mutation of the gene involved in MTOC orientation causes a neurological disorder, lissencephaly, that is associated with abnormal neuronal migration in the cerebral cortex (
      • Gleeson J.G.
      • Walsh C.A.
      ,
      • Gupta A.
      • Tsai L.H.
      • Wynshaw-Boris A.
      ). Therefore, the perinuclear localization of Cdc42 may reflect the importance of Cdc42 for the regulation of MTOC orientation in migrating neurons. Taken together, we propose a model of PI 3-kinase/Rac1- and Cdc42-linked radial migration of cortical neurons (Fig. 5). Although we could not rule out the involvement of other downstream effectors of PI 3-kinase in neuronal migration, our findings are consistent with previous studies on chemotaxis in non-neuronal cells and suggest similarities in cell migration between neuronal and non-neuronal cells. Our present study provides important insights into the molecular mechanisms underlying neuronal migration and cortical formation.
      Figure thumbnail gr5
      Fig. 5Model for the PI 3-kinase/Rac1 and Cdc42 signaling pathways in migrating neurons. PI 3-kinase triggered by extracellular stimuli including growth factors and/or extracellular matrix activates Rac1 and Cdc42 localized to the plasma membrane and peri-nuclear region containing the centrosome, respectively. Thus, the PI 3-kinase/Rac1 and Cdc42 pathways are synergistically involved in radial migration of cortical neurons. MT, microtubules.

      Acknowledgments

      We gratefully thank Drs. Shuh Narumiya, Toshiaki Katada, and Itaru Matsumura for the gift of plasmids. We also thank Takaki Miyata and Hidenori Tabata for the technical advice with slice culture and electroporation.

      References

        • Rice D.S.
        • Curran T.
        Annu. Rev. Neurosci. 2001; 24: 1005-1039
        • Gleeson J.G.
        • Walsh C.A.
        Trends Neurosci. 2000; 23: 352-359
        • Gupta A.
        • Tsai L.H.
        • Wynshaw-Boris A.
        Nat. Rev. Genet. 2002; 3: 342-355
        • Morris N.R.
        • Efimov V.P.
        • Xiang X.
        Trends Cell Biol. 1998; 8: 467-470
        • Xie Z.
        • Sanada K.
        • Samuels B.A.
        • Shih H.
        • Tsai L.H.
        Cell. 2003; 114: 469-482
        • Tanaka T.
        • Serneo F.F.
        • Tseng H.C.
        • Kulkarni A.B.
        • Tsai L.H.
        • Gleeson J.G.
        Neuron. 2004; 41: 215-227
        • Nikolic M.
        • Chou M.M.
        • Lu W.
        • Mayer B.J.
        • Tsai L.H.
        Nature. 1998; 395: 194-198
        • Iijima M.
        • Huang Y.E.
        • Devreotes P.
        Dev. Cell. 2002; 3: 469-478
        • Merlot S.
        • Firtel R.A.
        J. Cell Sci. 2003; 116: 3471-3478
        • Kawauchi T.
        • Chihama K.
        • Nabeshima Y.
        • Hoshino M.
        EMBO J. 2003; 22: 4190-4201
        • Niwa H.
        • Yamamura K.
        • Miyazaki J.
        Gene (Amst.). 1991; 108: 193-199
        • Ohkawa Y.
        • Hayashi K.
        • Sobue K.
        Biochem. Biophys. Res. Commun. 2003; 301: 78-83
        • Tabata H.
        • Nakajima K.
        J. Neurosci. Res. 2002; 69: 723-730
        • Hatanaka Y.
        • Murakami F.
        J. Comp. Neurol. 2002; 454: 1-14
        • Paik J.H.
        • Chae S.
        • Lee M.J.
        • Thangada S.
        • Hla T.
        J. Biol. Chem. 2001; 276: 11830-11837
        • Mercurio A.M.
        • Rabinovitz I.
        • Shaw L.M.
        Curr. Opin. Cell Biol. 2001; 13: 541-545
        • Fishman R.B.
        • Hatten M.E.
        J. Neurosci. 1993; 13: 3485-3495
        • Anton E.S.
        • Kreidberg J.A.
        • Rakic P.
        Neuron. 1999; 22: 277-289
        • Liesi P.
        EMBO J. 1985; 4: 1163-1170
        • Wang F.
        • Herzmark P.
        • Weiner O.D.
        • Srinivasan S.
        • Servant G.
        • Bourne H.R.
        Nat. Cell Biol. 2002; 4: 513-518
        • Weiner O.D.
        • Neilsen P.O.
        • Prestwich G.D.
        • Kirschner M.W.
        • Cantley L.C.
        • Bourne H.R.
        Nat. Cell Biol. 2002; 4: 509-513
        • Iijima M.
        • Devreotes P.
        Cell. 2002; 109: 599-610
        • Funamoto S.
        • Meili R.
        • Lee S.
        • Parry L.
        • Firtel R.A.
        Cell. 2002; 109: 611-623
        • Liu H.
        • Radisky D.C.
        • Wang F.
        • Bissell M.J.
        J. Cell Biol. 2004; 164: 603-612
        • Srinivasan S.
        • Wang F.
        • Glavas S.
        • Ott A.
        • Hofmann F.
        • Aktories K.
        • Kalman D.
        • Bourne H.R.
        J. Cell Biol. 2003; 160: 375-385
        • Keely P.J.
        • Westwick J.K.
        • Whitehead I.P.
        • Der C.J.
        • Parise L.V.
        Nature. 1997; 390: 632-636
        • Welch H.C.
        • Coadwell W.J.
        • Stephens L.R.
        • Hawkins P.T.
        FEBS Lett. 2003; 546: 93-97
        • Tolias K.F.
        • Cantley L.C.
        • Carpenter C.L.
        J. Biol. Chem. 1995; 270: 17656-17659
        • Zheng Y.
        • Bagrodia S.
        • Cerione R.A.
        J. Biol. Chem. 1994; 269: 18727-18730
        • del Pozo M.A.
        • Price L.S.
        • Alderson N.B.
        • Ren X.D.
        • Schwartz M.A.
        EMBO J. 2000; 19: 2008-2014
        • Palazzo A.F.
        • Joseph H.L.
        • Chen Y.J.
        • Dujardin D.L.
        • Alberts A.S.
        • Pfister K.K.
        • Vallee R.B.
        • Gundersen G.G.
        Curr. Biol. 2001; 11: 1536-1541
        • Etienne-Manneville S.
        • Hall A.
        Cell. 2001; 106: 489-498