Rescue of the Reeler Phenotype in theDentate Gyrus by Wild-Type Coculture Is

Mediated by Lipoprotein Receptors forReelin and Disabled 1

SHANTING ZHAO,1 XUEJUN CHAI,1 HANS H. BOCK,2 BIANKA BRUNNE,2

ECKART FORSTER,1AND MICHAEL FROTSCHER1,2*

1Institut fur Anatomie und Zellbiologie, Albert-Ludwigs-Universitat Freiburg, D-79104Freiburg, Germany

2Zentrum fur Neurowissenschaften, Albert-Ludwigs-Universitat Freiburg, D-79104Freiburg, Germany

ABSTRACTReelin is a positional signal for the lamination of the dentate gyrus. In the reeler mutant

lacking Reelin, granule cells are scattered all over the dentate gyrus. We have recently shownthat the reeler phenotype of the dentate gyrus can be rescued in vitro by coculturing reelerhippocampal slices with slices from wild-type hippocampus. Here we studied whether Reelinfrom other brain regions can similarly induce this rescue effect and whether it is mediated via theReelin receptors apolipoprotein E receptor 2 (ApoER2) and very-low-density lipoprotein receptor(VLDLR). We found that coculturing reeler hippocampal slices with slices from wild-type olfac-tory bulb, cerebellum, and neocortex rescued the reeler phenotype as seen before with hippocam-pal slices, provided that the Reelin-synthesizing cells of these regions were placed near themarginal zone of the reeler hippocampal slice. However, coculturing wild-type hippocampal sliceswith hippocampal slices from mutants deficient in ApoER2 and VLDLR did not rescue thereeler-like phenotype in these cultures. Similarly, no rescue of the reeler-like phenotype wasobserved in slices from mutants lacking Disabled 1 (Dab1), an adapter protein downstream ofReelin receptors. Conversely, reeler hippocampal slices were rescued by coculturing them withslices from Dab1�/� mutants or ApoER2�/�/VLDLR�/� mice. These findings show that Reelinfrom other brain regions can substitute for the loss of hippocampal Reelin and that rescue of thereeler phenotype observed in our coculture studies is mediated via lipoprotein receptors for Reelinand Dab1. J. Comp. Neurol. 495:1–9, 2006. © 2006 Wiley-Liss, Inc.

Indexing terms: reeler mouse; neuronal migration; Cajal-Retzius cells; ApoER2; VLDLR;

hippocampus

Little is known about the signals that control the char-acteristic formation of cell and fiber layers in the dentategyrus and the integration of postnatally generated gran-ule cells into the hippocampal network. The extracellularmatrix protein Reelin is known to play an important role,because the characteristic lamination of the dentate gyrusis lost in the reeler mutant lacking Reelin (Stanfield andCowan, 1979; Drakew et al., 2002). Reelin is synthesizedand secreted by Cajal-Retzius cells located in the outermolecular layer close to the hippocampal fissure(D’Arcangelo et al., 1995, 1997; del Rio et al., 1997;Frotscher, 1998), the marginal zone of the dentate gyrus.

We have recently shown that the reeler phenotype of thedentate gyrus can be rescued in slice cultures of reeler

hippocampus by coculturing them with wild-type hip-pocampal slices providing a Reelin-containing marginal

Grant sponsor: Deutsche Forschungsgemeinschaft; Grant number: SFB505; Grant number TR-3; Grant number BO 1806/2-1; Grant sponsor:Wolfgang Paul Program.

The first two authors contributed equally to this work.*Correspondence to: Michael Frotscher, Institut fur Anatomie und Zell-

biologie, Albert-Ludwigs-Universitat Freiburg, Albertstr. 17, D-79104Freiburg, Germany. E-mail: michael.frotscher@anat.uni-freiburg.de

Received 3 August 2005; Revised 2 September 2005; Accepted 20 Sep-tember 2005

DOI 10.1002/cne.20846Published online in Wiley InterScience (www.interscience.wiley.com).

THE JOURNAL OF COMPARATIVE NEUROLOGY 495:1–9 (2006)

© 2006 WILEY-LISS, INC.

zone (Zhao et al., 2004). After an incubation period ofabout 24 hours, the scattered granule cells of the reelerdentate gyrus formed a compact cell band reminiscent of agranular layer. In contrast, no granular layer formedwhen recombinant Reelin was added to the culture me-dium (Zhao et al., 2004). We concluded from these studiesthat Reelin has to be in a specific location to exert itseffects on granule cell lamination.

These experiments did not clarify the signaling cascadeinvolved. Several Reelin receptors have been described.The lipoprotein receptors apolipoprotein E receptor 2(ApoER2) and the very-low-density lipoprotein receptor(VLDLR) are known to be Reelin receptors (Trommsdorffet al., 1999). Alpha3beta1 integrins (Dulabon et al., 2000)have been reported to function as Reelin receptors as well,although binding of Reelin to them has recently beendisputed (Jossin et al., 2004). Disabled-1 (Dab1) acts as anadapter protein downstream to Reelin receptors and isphosphorylated upon Reelin binding (Hiesberger et al.,1999; Howell et al., 1999).

In the present study, we used our coculture assay todetermine the signaling molecules involved in the rescueof the reeler phenotype in the dentate gyrus. First, weshow that wild-type hippocampal slices can be replaced byslices of neocortex, olfactory bulb, and cerebellum, whichall contain different neuronal types synthesizing Reelin,suggesting that the rescue effect is brought about by Ree-lin and does not depend on tissue-inherent properties. Wethen provide evidence that Reelin signaling via ApoER2and VLDLR and the adapter protein Dab1 underlies theobserved rescue of granule cell lamination in slices ofreeler hippocampus.

MATERIALS AND METHODS

Preparation of slice cultures

For the preparation of slice cultures, newborn (P0)homozygous reeler mice, newborn (P0) and young post-natal (P3-5) ApoER2/VLDLR double knockout mice,Dab1 knockout mice, and wild-type mice or rats wereused. The mutant mice were identified by their well-known morphological malformations in the cortex andhippocampus; the genotype of reeler mice, ApoER2/VLDLR double knockout mice, and Dab1 mutants wasconfirmed by PCR analysis of genomic DNAs as de-scribed elsewhere (Deller et al., 1999; Beffert et al.,2002). Brains were removed following decapitation un-der hypothermic anesthesia, and the hippocampi, neo-cortex, olfactory bulb, and cerebellum were dissectedand sliced (300 �m) with a Mcllwain tissue chopper. Allexperiments were performed in agreement with the in-stitutional guide for animal care.

The following cocultures were prepared (at least 10 co-cultures per experimental group): hippocampal slices fromP0 reeler mice cocultured to slices of neocortex from P0wild-type mice or to slices of hippocampus, olfactory bulb,and cerebellum from P3–5 wild-type mice or rats; hip-pocampal slices from P0 reeler mice to hippocampal slicesfrom ApoER2/VLDLR double knockout mice or Dab1 mu-tant mice (P5); and hippocampal slices from ApoER2/VLDLR double knockout mice or Dab1 mutants (P0) tohippocampal slices from P3–5 wild-type mice or rats. Inthe cocultures of reeler hippocampus and wild-type neo-cortex, the reeler dentate gyrus was placed next to the

Reelin-containing marginal zone of neocortex or, alterna-tively, to the Reelin-negative ventricular zone. In the co-cultures of reeler hippocampus and wild-type olfactorybulb, the lamina glomerulosa of olfactory bulb was re-moved with a scalpel to expose the Reelin-synthesizingmitral cells. Slices were transferred onto Millipore mem-branes, which were placed in a six-well plate with 1 ml/well nutrition medium (25% heat-inactivated horse se-rum, 25% Hank�s balanced salt solution, 50% minimalessential medium, 2 mM glutamine, pH 7.2) before use.Slices were incubated as static cultures in 5% CO2 at 37°Cfor 7–10 days in vitro (DIV; Stoppini et al., 1991). Themedium was changed every 2 days.

Immunocytochemistry

For Reelin immunostaining of newborn and young post-natal mice, brains of wild-type animals were removedfollowing decapitation and fixed by immersion in 4% para-formaldehyde in 0.1 M phosphate buffer (PB; pH 7.4)overnight. Then, the brains were cut on a vibratome (50�m).

Cultures were fixed after 7–10 DIV with 4% paraformal-dehyde for 2 hours and then washed several times in 0.1 MPB. Thereafter, the cultures were resliced on a vibratome(50 �m). Tissue sections and sections of cultures werepreincubated in a blocking solution (5% normal goat se-rum, 0.2% Triton-X 100 in 0.1 M PB) for 30 minutes atroom temperature. After being rinsed in 0.1 M PB, sec-tions were incubated with the following primary antibod-ies overnight at 4°C: monoclonal mouse anti-Reelin G10(1:1,000; Chemicon, Hofheim, Germany; MAB 5364, im-munogen: a fusion protein containing amino acids 164–496 of mouse Reelin) or polyclonal rabbit anti-Prox-1 (1:1,000; Chemicon; AB 5475, immunogen: last 15 aminoacids of the C-terminus of mouse Prox-1; see Bagri et al.,2002). After being washed in 0.1 M PB, sections wereincubated in secondary antibodies (Alexa 488 goat anti-rabbit IgG, A11008, and/or Alexa 568 goat anti-mouseIgG, A11004; 1:600; Molecular Probes, Gottingen, Ger-many) at 4°C overnight. The monoclonal antibody G10against Reelin specifically stained Reelin-synthesizingneurons as originally described by de Bergeyck et al.(1998; see Fig. 1). It stained the 400-kD Reelin protein andthe 300-kD and 180-kD Reelin fragments in Western blots(see Fig. 2). Identical immunolabeling was observed withthe original G10 antibody (de Bergeyck et al., 1998; kindlyprovided by Dr. A. Goffinet) and the commercial G10 an-tibody from Chemicon. No immunostaining was observedwith tissue from reeler mutants lacking Reelin. TheProx-1 antibody specifically stained the granule cells ofthe dentate gyrus as previously described (Bagri et al.,2002). No double labeling was observed with markers forGABAergic interneurons and mossy cells (data notshown). In the present experiment, we used this charac-teristic staining pattern of Prox-1 to monitor the rescue ofgranule cell lamination by wild-type coculture (see, e.g.,Fig. 3).

Sections of immersion-fixed brains were incubated withneurotrace (1:500; Molecular Probes; N21480) for 30 min-utes at room temperature. After being rinsed in 0.1 M PBfor 2 hours, they were mounted in Moviol.

Sections were photographed with a Zeiss LSM-510 con-focal microscope. Confocal images were exported from theZeiss LSM image browser and stored as TIFF files. Fig-ures were prepared in Adobe PhotoShop 6.0 and Adobe

2 S. ZHAO ET AL.

Illustrator 10 (Adobe, San Jose, CA). Image brightness,contrast, and sharpness were adjusted.

Western blot analysis for Reelin

Western blot analysis included lysates of hippocampaltissues from P0 reeler mice and P0 rats, lysates of neocor-tex from P0 wild-type mouse, and olfactory bulb and cer-ebellum from P3 wild-type mice. Brains were removedfollowing decapitation under hypothermic anesthesia. Thehippocampus, neocortex, olfactory bulb, and cerebellumwere disected and collected in 1.5-ml tubes on ice. Thesamples were weighed and immediately frozen in liquidnitrogen. Six volumes (v/w) of hypotonic lysis buffer (50mM Tris buffer, 150 mM NaCl, 5 mM EDTA, and 1%protease inhibitor cocktail; Sigma, Munich, Germany; pH7.6) were added to each sample, and the tissue was lysedby repeated thawing at 37°C and freezing in liquid nitro-gen (five times). After sonication for 5 minutes and tritu-ration with a pipette tip to homogenize larger tissuepieces, the suspension was centrifuged at 20,000g for 20minutes (0°C). The resulting crude supernatants werestored at –80°C or directly used. Supernatants (20 �g)from freshly prepared tissues were diluted with samplebuffer (Invitrogen, Karlsruhe, Germany) and boiled for 5minutes. Proteins were separated by 3–8% gradient Tris-acetate gel electrophoresis (SDS-PAGE; Invitrogen) andtransferred electrophoretically to polyvinylidene fluoride(PVDF) membranes. A monoclonal mouse antibodyagainst Reelin G10 (Chemicon; MAB 5364, immunogen:recombinant Reelin amino acids 164–496) was used asprimary antibody at a dilution of 1:3,000, followed by analkaline phosphatase-conjugated secondary antibody(goat anti-mouse IgG, 1:10,000; Invitrogen; G21060). Theimmunoreaction was visualized by a chemiluminescencereaction (Invitrogen).

RESULTS

Reelin expressed by different cell types indifferent brain regions is qualitatively

similar

Reelin is synthesized in different brain regions and bydifferent types of cells. In the hippocampus and neocortex,Cajal-Retzius (CR) cells in the marginal zone synthesizeReelin. Thus, CR cells of the dentate gyrus are concen-trated in the outer molecular layer, the marginal zone ofthe dentate gyrus (Fig. 1A). The granule cells, nonimmu-noreactive to Reelin, accumulate underneath, therebyforming a second c-shaped band beneath the larger, sim-ilarly c-shaped layer of CR cells. Neocortical sections fromP0 mice show a distinct labeling of CR cells in the mar-ginal zone, future layer I, when immunostained for Reelin(Fig. 1B). Nonimmunoreactive neurons seem to arrest un-derneath, suggesting that they were prevented from in-vading the marginal zone. In the olfactory bulb, Reelin issynthesized by the mitral cells, which form a ring underwhich nonimmunoreactive neurons accumulate (Fig. 1C).In the cerebellum, Reelin is known to be synthesized bygranule cell precursors in the external granular layer (Fig.1D). The size and ratio of Reelin and its fragments fromthese different cell types and brain regions were similarwhen analyzed by Western blotting (Fig. 2).

Rescue of the reeler phenotype by Reelinfrom different sources

As described previously, coculturing of a reeler slicenext to a wild-type slice with the Reelin-containing mar-ginal zone of the wild-type dentate gyrus attached to thereeler dentate gyrus induced formation of a compact gran-ular layer in the reeler slice (Zhao et al., 2004; Fig. 3A).Because Reelin from different brain regions and cell typesturned out to be similar (Fig. 2), we reasoned that it mightbe possible to rescue the reeler phenotype of the dentategyrus by coculturing reeler slices with wild-type tissuefrom these other regions. In fact, when the marginal zoneof a neocortical slice was placed next to the dentate gyrusof a reeler slice, we again observed the formation of acompact granular layer (Fig. 3B). Similarly, a dense bandof accumulating granule cells was observed when a reelerslice was placed next to Reelin-synthesizing mitral cells ofthe olfactory bulb (Fig. 3C). Finally, a granular layer alsoformed when the reeler tissue was cocultured with a cer-ebellar slice containing Reelin-synthesizing cerebellargranule cell precursors (Fig. 3D). These findings indicatethat Reelin from different cell types and brain regions canrescue the reeler phenotype of the dentate gyrus. More-over, these experiments show that tissue-specific proper-ties or factors depending on the developmental stage of thesecreting cells are not critical for the observed Reelineffect.

The Reelin-synthesizing cells of the various brain re-gions had to be in close apposition to the reeler dentategyrus to provide a Reelin-containing marginal zone. Thus,no rescue was seen when the ventricular zone of a neocor-tical slice was cocultured next to the reeler dentate gyrus.Accordingly, when a neocortical slice was cocultured withtwo reeler slices, one adjacent to the marginal zone andthe other one next to the ventricular zone (Fig. 4A), onlythe reeler slice opposed to the marginal zone showed theformation of a compact granular layer (Fig. 4B,C).

Lipoprotein receptors for Reelin and Dab1are required for granule cell lamination

Double-knockout mice deficient in ApoER2 and VLDLRshow a phenotype indistinguishable from that of reeler(Trommsdorff et al., 1999; Drakew et al., 2002). We ac-cordingly hypothesized that the rescue of granule celllamination in slices from postnatal reeler mutants ob-served after coculturing with wild-type is mediated bylipoprotein receptors for Reelin. In fact, coculturing ofwild-type tissue with ApoER2�/�/VLDLR�/� mutantslices, which like wild-type slices expressed Reelin butotherwise showed the scattered distribution of granulecells characteristic of the reeler phenotype, did not inducethe formation of a compact granule cell layer in the recep-tor mutant slices (Fig. 5A). Similarly, no rescue of granulecell lamination was observed when wild-type slices werecocultured with slices from Dab1 knockout mice (Fig. 5B).Dab1 mutants, as with ApoER2/VLDLR double-knockoutmice, exhibit a reeler-like phenotype. Conversely, bothslices from ApoER2�/�/VLDLR�/� mutants and slicesfrom Dab1-deficient mice rescued the reeler phenotype incocultured reeler slices, likely because of their Reelin ex-pression (Fig. 6A,B). Together these findings show thatlipoprotein receptors for Reelin and the adapter proteinDab1 are required for the rescue of granule cell laminationin slices of reeler hippocampus. Because we used slices

3LAYER FORMATION IN THE DENTATE GYRUS

from postnatal hippocampus, our results also suggest thatthese molecules control the integration of postnatally gen-erated granule cells into the granular layer in vivo.

DISCUSSION

By coculturing reeler slices and wild-type slices with themarginal zone of the wild-type slice opposing the reelerdentate gyrus, we could rescue the reeler phenotype, thusconfirming the results of our previous studies (Zhao et al.,2004). Here we provide evidence that Reelin from differentbrain regions and different cell types can exert this effect.Moreover, by coculturing wild-type tissue to slices fromReelin receptor mutants and mutants lacking the adapterprotein Disabled-1, we provide evidence that the rescue ofgranule cell lamination is mediated via these molecules ofthe Reelin signaling cascade. We conclude that the forma-tion of a laminated dentate gyrus and the integration of

Fig. 1. Reelin is expressed in different brain regions and by dif-ferent cell types. A: Immunostaining of mouse hippocampus (P3) forReelin (red) labels Cajal-Retzius cells in the marginal zone of thedentate gyrus, the outer molecular layer (oml). B: Reelin immuno-staining of Cajal-Retzius cells in the marginal zone (MZ) of the mouseneocortex (P0). C: Reelin-immunoreactive mitral cells (MC) of the P3mouse olfactory bulb. D: Reelin-immunoreactive cells in the external

granular layer (EGL) of the P3 mouse cerebellum. In A–C, nonimmu-noreactive neurons are counterstained with Neurotrace (green). CP,cortical plate; DG, dentate gyrus; EGL, external granular layer; g,granular layer; h, hilus; iml, inner molecular layer; IZ, intermediatezone; MC, mitral cells; MZ, marginal zone; oml, outer molecular layer;SP, subplate. Scale bars � 80 �m in A,B; 250 �m in C,D.

Fig. 2. Western blots for Reelin from various brain regions arequalitatively similar. Lane 1: P3 wild-type hippocampus; lane 2: P0wild-type neocortex; lane 3: P3 wild-type olfactory bulb; lane 4: P3wild-type cerebellum; lane 5: P3 reeler hippocampus.

4 S. ZHAO ET AL.

postnatally generated granule cells require Reelin signal-ing via ApoER2, VLDLR, and Dab1.

Reelin from different sources is equivalentin rescuing granule cell lamination

Developmental processes in the various brain regionscontaining Reelin-synthesizing cells are quite different,

the only common trait being the formation of laminatedstructures. In the neocortex, Reelin controls the formationof cortical layers in an inside-out manner, with late-generated superficial neurons migrating past the early-generated neurons of the deep layers. In contrast, in thedentate gyrus, Reelin is required for the development of atightly packed granular layer, which develops in an

Fig. 3. Rescue of granule cell lamination in the reeler dentategyrus is achieved by coculturing reeler slices with slices from differentbrain regions containing Reelin-immunoreactive cells (red). The for-mation of a compact layer of granule cells (immunostained for Prox-1,green) is seen when the Reelin-synthesizing cells are cocultured inclose proximity to the reeler dentate gyrus. Dotted lines indicateborders between cultures. A: Coculture of P0 reeler hippocampus and

P3 wild-type hippocampus. B: P0 reeler hippocampus cocultured to P3wild-type neocortex. Reelin-immunoreactive Cajal-Retzius cells in themarginal zone (MZ) are directly attached to the reeler culture. C: P0reeler hippocampus cocultured to mitral cells (MC) from P3 wild-typeolfactory bulb. D: P0 reeler hippocampus cocultured to P3 wild-typecerebellum. Abbreviations as in Figure 1. Scale bars � 100 �m.

5LAYER FORMATION IN THE DENTATE GYRUS

Fig. 4. Positioning of the marginal zone containing Reelin-immunoreactive Cajal-Retzius cells is crucial for the rescue of granulecell lamination. A: DAPI-stained triplet culture with a wild-type neo-cortical slice (P0) sandwiched between two reeler cultures (P0). Atightly packed granule cell layer (g) has formed only in the reelerculture (DG1) facing the marginal zone (MZ) of the neocortical cul-ture. In the reeler culture (DG2) facing the ventricular zone (VZ),neurons remain loosely distributed all over the dentate gyrus. Dottedlines indicate borders between cultures; boxed areas are shown in B

and C, respectively. B: Reeler culture (DG1) placed next to the mar-ginal zone of wild-type neocortex shown at higher magnification.Double labeling for Prox-1 to label granule cells (green) and Reelin(red) to label CR cells. Note the formation of a dense layer of granulecells (g) near Reelin-immunoreactive CR cells. C: Reeler culture(DG2) placed next to the ventricular zone (VZ) of the wild-type cul-ture, remote from Reelin-synthesizing CR cells. No compact granulecell layer has formed. Scale bars � 200 �m in A; 60 �m in B,C.

outside-in fashion. In the olfactory bulb, Reelin was foundto act as a detachment signal for chain-migrating precur-sors of olfactory bulb interneurons (Hack et al., 2002). Inthe cerebellum, Reelin synthesized by granule cell precur-sors in the external granular layer controls the properpositioning of Purkinje cells (Miyata et al., 1997). Despitethese various functions in brain development, Reelin fromthese different brain regions invariably rescued the for-mation of a compact granule cell layer in our cocultureexperiments. In line with this uniform effect of Reelinfrom different sources, Western blots for Reelin from theseregions gave very similar results. We conclude that devel-opmental differences among these brain regions arisefrom different Reelin receptors of the target cells, differ-ences in the Reelin signaling pathway, and/or additionalmolecular signals.

ApoER2, VLDLR, and Dab1 controldevelopment of the dentate gyrus

Several Reelin receptors have been described. There isreason to assume that, for development of the dentategyrus and for integration of postnatally generated granulecells, the lipoprotein receptors ApoER2 and VLDLR aswell as the adapter protein Disabled-1 play an importantrole. In the absence of lipoprotein receptors or Dab1, nocompact granule cell layer formed in coculture experi-ments. These findings do not exclude a role for otherReelin receptors, such as alpha3beta1 integrins. In fact,localized migration defects of the granule cells, likelycaused by a maldifferentiation of the radial glial scaffold,have been described for alpha3beta1-deficient mutants

(Forster et al., 2002). In comparing ApoER2 mutant miceand mice lacking VLDLR, the former were found to showa more severe migration defect in the dentate gyrus(Drakew et al., 2002). Absence of ApoER2 plus VLDLRresults in a phenotype indistinguishable from that ofreeler mutants (Trommsdorff et al., 1999; Drakew et al.,2002). Reelin binding to both types of lipoprotein receptorsinvolves Dab1, and scrambler mutants lacking Dab1 ex-hibit a reeler phenotype (Howell et al., 1997; Sheldon etal., 1997; Ware et al., 1997; Hiesberger et al., 1999). Theresults of the present study add to the observations madein these mutants that Reelin, its lipoprotein receptors,and Dab1 have to be present as late as in the early post-natal period to allow for the formation of a compact gran-ule cell layer in the dentate gyrus.

Reelin binding to lipoprotein receptors induces Dab1phosphorylation by the nonreceptor tyrosine kinase fyn(Arnaud et al., 2003; Bock and Herz, 2003) and modulatestau phosphorylation (Hiesberger et al., 1999; Beffert et al.,2002). Phosphorylation of Dab1 at tyrosines 220 and 232was recently found to be involved in the disconnection ofthe migrating neuron from the radial glial fiber, whichterminates the migration process (Sanada et al., 2004).Thus, phosphorylation of Dab1 may underlie the pre-sumed stop signal function of Reelin (Curran andD’Arcangelo, 1998; Frotscher, 1998). In the rescue exper-iments of the present study, we regularly observed anaccumulation of granule cells at some distance from theReelin-synthesizing Cajal-Retzius cells of the coculturedwild-type tissue. Similarly, in the wild-type dentate gyrus,granule cells form a densely packed layer at some distance

Fig. 5. Rescue of granule cell lamination does not occur in culturesfrom ApoER2�/�/VLDLR�/� cultures and cultures lacking Dab1.A: Coculture of an ApoER2�/�/VLDLR�/� culture (P0) with a cultureof P3 wild-type hippocampus. No formation of a compact granule celllayer is seen in the mutant culture. Note the presence of Reelinimmunoreactivity in the marginal zone of the mutant culture. Den-

tate granule cells immunlabeled for Prox-1 (green). B: No rescue ofgranule cell lamination is seen when a Dab1�/� slice (P0) was cocul-tured with a P3 wild-type hippocampal slice. Prox-1-immunoreactivegranule cells in the mutant slice are scattered all over the dentategyrus. Note Reelin-immunoreactive Cajal-Retzius cells in the mar-ginal zone of the slice from the Dab1�/� mutant. Scale bars � 100 �m.

7LAYER FORMATION IN THE DENTATE GYRUS

from the CR cells in the marginal zone and do not invadethe molecular layer.

Role of Reelin in laminating thedentate gyrus

Whereas various molecules of the Reelin signaling cas-cade have been identified in recent years, we still knowvery little about the role of Reelin signaling in the varioussteps leading to the formation of the granular layer. Asmentioned, Reelin has been regarded as a stop signal formigrating neurons brought about by Dab1 phosphoryla-tion resulting in the disconnection of the migrating neuronfrom the radial glial fiber (Sanada et al., 2004). In thereeler mutant, many granule cells remain in the hilus anddo not migrate toward the granular layer, suggesting thatReelin in the marginal zone could also act as an attractivesignal. It should be noted, however, that Reelin, in addi-tion to its direct effects on neurons, also acts on glialfibrillary acidic protein (GFAP)-positive radial glial cellsby increasing their process length (Forster et al., 2002;Frotscher et al., 2003; Hartfuss et al., 2003; Weiss et al.,2003). Moreover, Reelin was found to be a positional sig-nal for outgrowing radial glial fibers (Zhao et al., 2004).We hypothesize, based on these data, a dual function ofReelin in the development of the dentate gyrus. As a firststep, Reelin induces the formation of a regular radial glialscaffold with the outgrowing radial fibers heading towardthe Reelin-containing marginal zone. Granule cells thenmigrate along these radial glial fibers to reach their finaldestination in the granular layer. This, in turn, is

achieved by Reelin’s second function, its direct effect onthe migrating neuron. After Reelin binding to neuronallipoprotein receptors, Dab1 is phosphorylated, which re-sults in the disconnection of the neuron from the radialglial fiber. This terminates neuronal migration and leadsto an accumulation of granule cells in the granular layer(Zhao et al., 2004).

Alternatively, radial glial fibers extending toward theReelin-containing marginal zone may be inherited bynewly generated granule cells and become their apicaldendrites (Miyata et al., 2001), insofar as radial glial cellsare precursors of neurons (Malatesta et al., 2000; Noctoret al., 2001). Migration of the granule cell from the prolif-eration zone in the hilus may take place by nuclear trans-location (Kriegstein and Noctor, 2004), a process that maybe terminated by Reelin signaling. Future live microscopystudies will be needed to determine which of the twoscenarios predominates during the long-lasting period ofgranule cell neurogenesis or holds true at all for the de-velopment of the dentate gyrus.

ACKNOWLEDGMENTS

We thank Cornelia Hofmann and Claudia Kobele fortheir excellent technical assistance.

LITERATURE CITED

Arnaud L, Ballif BA, Forster E, Cooper JA. 2003. Fyn tyrosine kinase is acritical regulator of disabled-1 during brain development. Curr Biol13:9–17.

Fig. 6. ApoER2�/�/VLDLR�/� slices and slices from Dab1�/� micerescue granule cell lamination in the reeler dentate gyrus. A: Cocul-ture of a P0 reeler hippocampal slice with a hippocampal slice from anApoER2�/�/VLDLR�/� mutant. The border between the two culturesis indicated by a dotted line. Note the formation of a tightly packedgranular layer in the reeler slice facing the Reelin-containing mar-ginal zone (red) of the ApoER2�/�/VLDLR�/� slice. Note scattered

Prox-1-positive granule cells (green) in the slice from the lipoproteinreceptor mutant. B: Formation of a Prox-1-positive, tightly packedgranule cell layer in a P0 reeler hippocampal slice cocultured to aDab1�/� hippocampal slice with Reelin-immunoreactive Cajal-Retzius cells (red) in its marginal zone. Note scattered distribution ofProx-1-positive granule cells in the dentate gyrus of the Dab1�/� slice.Scale bars � 100 �m.

8 S. ZHAO ET AL.

Bagri A, Gurney T, He X, Zou Y-R, Littman DR, Tessier-Lavigne M,Pleasure SJ. 2002. The chemokine SDF1 regulates migration of den-tate granule cells. Development 129:4249–4260.

Beffert U, Morfini G, Bock HH, Reyna H, Brady ST, Herz J. 2002. Reelin-mediated signaling locally regulates protein kinase B/Akt and glycogensynthase kinase 3beta. J Biol Chem 277:49958–49964.

Bock HH, Herz J. 2003. Reelin activates Src family tyrosine kinases inneurons. Curr Biol 13:18–26.

Curran T, D’Arcangelo G. 1998. Role of reelin in the control of braindevelopment. Brain Res Rev 26:285–294.

D’Arcangelo G, Miao GG, Chen S-C, Soares HD, Morgan JI, Curran T.1995. A protein related to extracellular matrix proteins deleted in themouse mutant reeler. Nature 374:719–723.

D’Arcangelo G, Nakajima K, Miyata T, Ogawa M, Mikoshiba K, Curran T.1997. Reelin is a secreted glycoprotein recognized by the CR-50 mono-clonal antibody. J Neurosci 17:23–31.

de Bergeyck V, Naerhuyzen B, Goffinet AM, Lambert de Rouvroit C. 1998.A panel of monoclonal antibodies against reelin, the extracellular ma-trix protein defective in reeler mutant mice. J Neurosci Methods 82:17–24.

Del Rio JA, Heimrich B, Borrell V, Forster E, Drakew A, Alcantara S,Nakajima K, Miyata T, Ogawa M, Mikoshiba K, Derer P, Frotscher M,Soriano E. 1997. A role for Cajal-Retzius cells and reelin in the devel-opment of hippocampal connections. Nature 385:70–74.

Deller T, Drakew A, Heimrich B, Forster E, Tielsch A, Frotscher M. 1999.The hippocampus of the reeler mutant mouse: fiber segregation in areaCA1 depends on the position of the postsynaptic target cells. ExpNeurol 156:254–267.

Drakew A, Deller T, Heimrich B, Gebhardt C, del Turco D, Tielsch A,Forster E, Herz J, Frotscher M. 2002. Dentate granule cells in reelermutants and VLDLR and ApoER2 knockout mice. Exp Neurol 176:12–24.

Dulabon L, Olson EC, Taglienti MG, Eisenhuth, S, McGrath B, Walsh CA,Kreidberg JA, Anton ES. 2000. Reelin binds alph3beta1 integrin andinhibits neuronal migration. Neuron 27:33–44.

Forster E, Tielsch A, Saum B, Weiss KH, Johanssen C, Graus-Porta D,Muller U, Frotscher M. 2002. Reelin, Disabled 1, and beta 1 integrinsare required for the formation of the radial glial scaffold in the hip-pocampus. Proc Natl Acad Sci U S A 99:13178–13183.

Frotscher M. 1998. Cajal-Retzius cells, reelin, and the formation of layers.Curr Opin Neurobiol 8:570–575.

Frotscher M, Haas CA, Forster E. 2003. Reelin controls granule cell mi-gration in the dentate gyrus by acting on the radial glial scaffold. CerebCortex 13:634–640.

Hack I, Bancila M, Loulier K, Carroll P, Cremer H. 2002. Reelin is adetachment signal in tangential chain-migration during postnatal neu-rogenesis. Nat Neurosci 5:939–945.

Hartfuss E, Forster E, Bock HH, Hack MA, Leprince P, Luque JM, Herz J,Frotscher M, Gotz M. 2003. Reelin signaling directly affects radial gliamorphology and biochemical maturation. Development 130:4597–4609.

Hiesberger T, Trommsdorff M, Howell BW, Goffinet A, Mumby MC, Cooper

JA, Herz J. 1999. Direct binding of Reelin to VLDL receptor and ApoEreceptor 2 induces tyrosine phosphorylation of disabled-1 and modu-lates tau phosphorylation. Neuron 24:481–489.

Howell BW, Hawkes R, Soriano P, Cooper JA. 1997. Neuronal position inthe developing brain is regulated by mouse disabled-1. Nature 389:733–737.

Howell BW, Herrick TM, Cooper JA. 1999. Reelin-induced tyrosine phos-phorylation of disabled 1 during neuronal positioning. Genes Dev 13:643–648.

Jossin Y, Ignatova N, Hiesberger T, Herz J, Lambert de Rouvroit C,Goffinet AM. 2004. The central fragment of Reelin, generated by pro-teolytic processing in vivo, is critical to its function during cortical platedevelopment. J Neurosci 24:514–521.

Kriegstein AR, Noctor SC. 2004. Patterns of neuronal migration in theembryonic cortex. Trends Neurosci 27:392–399.

Malatesta P, Hartfuss E, Gotz M. 2000. Isolation of radial glial cells byfluorescent-activated cell sorting reveals a neuronal lineage. Develop-ment 127:5253–5263.

Miyata T, Nakajima K, Mikoshiba K, Ogawa M. 1997. Regulation of Pur-kinje cell alignment by reelin as revealed with CR-50 antibody. J Neu-rosci 17:3599–3609.

Miyata T, Kawaguchi A, Okano H, Ogawa M. 2001. Asymmetric inheri-tance of radial glial fibers by cortical neurons. Neuron 31:727–741.

Noctor SC, Flint AC, Weissman TA, Dammermann RS, Kriegstein AR.2001. Neurons derived from radial glial cells establish radial units inneocortex. Nature 409:714–720.

Sanada K, Gupta A, Tsai L-H. 2004. Disabled-1-regulated adhesion ofmigrating neurons to radial glial fiber contributes to neuronal position-ing during early corticogenesis. Neuron 42:197–211.

Sheldon M, Rice DS, D’Arcangelo G, Yoneshima H, Nakajima K, MikoshibaK, Howell BW, Cooper JA, Goldowitz D, Curran T. 1997. Scrambler andyotari disrupt the disabled gene and produce a reeler-like phenotype inmice. Nature 389:730–733.

Stanfield BB, Cowan WM. 1979. The morphology of the hippocampus anddentate gyrus in normal and reeler mice. J Comp Neurol 185:393–422.

Stoppini L, Buchs PA, Muller D. 1991. A simple method for organotypiccultures of nervous tissue. J Neurosci Methods 37:173–182.

Trommsdorff M, Gotthardt M, Hiesberger T, Shelton J, Stockinger W,Nimpf J, Hammer RE, Richardson JA, Herz J. 1999. Reeler/disabled-like disruption of neuronal migration in knockout mice lacking theVLDL receptor and ApoE receptor 2. Cell 97:689–701.

Ware ML, Fox JW, Gonzalez JL, Davis NM, Lambert de Rouvroit C, RussoCJ, Chua SC Jr, Goffinet AM, Walsh CA. 1997. Aberrant splicing of amouse disabled homolog, mdab1, in the scrambler mouse. Neuron19:239–249.

Weiss KH, Johanssen C, Tielsch A, Herz J, Deller T, Frotscher M, ForsterE. 2003. Malformation of the radial glial scaffold in the dentate gyrusof reeler mice, scrambler mice, and ApoER2/VLDLR-deficient mice.J Comp Neurol 460:56–65.

Zhao S, Chai X, Forster E, Frotscher M. 2004. Reelin is a positional signalfor the lamination of dentate granule cells. Development 131:5117–5125.

9LAYER FORMATION IN THE DENTATE GYRUS