Development 126, 5285-5294 (1999) 5285 Printed in Great Britain © The Company of Biologists Limited 1999 DEV1475

Ebf1 controls early cell differentiation in the embryonic striatum

Sonia Garel1, Faustino Marín1, Rudolf Grosschedl2,* and Patrick Charnay1,‡ 1Unité 368 de l’Institut National de la Santé et de la Recherche Médicale, Ecole Normale Supérieure, 46 rue d’Ulm, 75230 Paris Cedex 05, France 2Howard Hughes Medical Institute, Department of Microbiology and Immunology, University of California, San Francisco, California 94143, USA *Present address: Institute for Biochemistry and Centre, University of Munich, Feodor-Lynen Str. 25, 81377 München, Germany ‡Author for correspondence (e-mail: [email protected])

Accepted 22 September; published on WWW 9 November 1999

SUMMARY

Ebf1/Olf-1 belongs to a small multigene family encoding adhesion/guidance molecules. These early defects in the closely related helix-loop-helix transcription factors, SVZ/mantle transition are followed by an increase in cell which have been proposed to play a role in neuronal death, a dramatic reduction in size of the postnatal differentiation. Here we show that Ebf1 controls cell striatum and defects in navigation and fasciculation of differentiation in the murine embryonic striatum, where it thalamocortical fibres travelling through the striatum. Our is the only gene of the family to be expressed. Ebf1 targeted data therefore show that Ebf1 plays an essential role in the disruption affects postmitotic cells that leave the acquisition of mantle cell molecular identity in the subventricular zone (SVZ) en route to the mantle: they developing striatum and provide information on the genetic appear to be unable to downregulate normally hierarchies that govern neuronal differentiation in the restricted to the SVZ or to activate some mantle-specific ventral telencephalon. genes. These downstream genes a variety of regulatory including transcription factors and Key words: Ebf, Olf, Striatum, Neurogenesis, Apoptosis, Axonal proteins involved in retinoid signalling as well as navigation

INTRODUCTION non-basic HLH dimerisation domain (Hagman et al., 1995). Two closely related mouse genes were subsequently identified, The vertebrate central nervous system (CNS) is a highly Ebf2 (also known as O/E-3 or Mmot1) and Ebf3 (also known complex structure whose normal activity relies on the as O/E-2) (Garel et al., 1997; Malgaretti et al., 1997; Wang et appropriate differentiation and interconnection of a variety of al., 1997). In the embryonic CNS, the three genes show neuronal subtypes. The control of neuronal differentiation overlapping expression patterns with two salient features: (1) therefore constitutes a key aspect of CNS development. In the they are transiently expressed in differentiating neurons, Ebf2 last few years a number of vertebrate genes involved in this being expressed in early postmitotic neurons and Ebf1 and process have been identified on the basis of homology with Ebf3 in more differentiated cells; (2) they are expressed along Drosophila genes. Hence, genes encoding members of the basic the entire rostrocaudal axis from the midbrain to the spinal helix-loop-helix (bHLH) family of transcription factors like cord, whereas in the forebrain they are regionally restricted Mash-1 and the have been implicated in various (Garel et al., 1997). These results suggested a role for Ebf steps in the determination and differentiation of the CNS and genes in neuronal differentiation and specification of forebrain of the peripheral nervous system (PNS) (Guillemot et al., 1993; subpopulations, which is supported by experiments in Xenopus Cau et al., 1997; Fode et al., 1998; Hirsch et al., 1998; Ma et and C. elegans embryos: alteration of the Ebf2 ortholog al., 1998; Casarosa et al., 1999). Nevertheless our understanding (XCoe2) activity in Xenopus embryos revealed a role for this of the genetic networks governing neuronal differentiation is factor in early steps of primary neuron differentiation (Dubois still very fragmentary. In the present paper, we provide direct et al., 1998), and mutations in the C. elegans Ebf ortholog, unc- evidence of the involvement of another gene family, the Ebf/Olf 3, affect motor neuron differentiation and ventral nerve cord family (also called O/E, Coe), in this process. fasciculation (Prasad et al., 1998). However, the role of Ebf The founding member of the family, Ebf1/Olf-1, was genes in mouse CNS development remains to be elucidated. originally cloned independently in mouse and rat on the basis Ebf1-deficient mice have been generated and show an early of its putative role in B-cell and olfactory neurons block in the differentiation of the B-cell lineage at the pro-B differentiation, respectively (Hagman et al., 1991, 1993; Wang stage (Lin and Grosschedl, 1995). In contrast, no abnormality and Reed, 1993). Ebf1 encodes a in the olfactory epithelium and no global defect in the containing an atypical zinc finger DNA binding domain and a CNS were observed in these mice, presumably because of 5286 S. Garel and others complementation or redundancy between the different Ebf fixed by immersion or perfused through the heart with 4% genes. However, since Ebf1 is the only member of the family paraformaldehyde (PFA) in PBS. Brains were cryoprotected in 30% expressed in the developing striatum (Garel et al., 1997), this sucrose in PBS and cut into 20-30 µm sections on a cryotome or area of the telencephalon deserved further analysis. embedded in paraffin and cut into 10 µm sections on a microtome. The striatum is a large structure located in the ventral Sections were preincubated in 5% foetal calf serum, 0.1% Tween-20 in telencephalon. During embryogenesis, the telencephalic vesicles PBS for 1 hour at room temperature (RT) and incubated overnight at 4¡C with: mouse anti-TH, 1:200 dilution (Boehringer Mannheim); differentiate dorsally into the cortex and ventrally into two bulges, mouse anti-DARPP-32, 1:5000 (a kind gift of J.A. Girault); rabbit anti- the lateral (LGE) and medial (MGE) ganglionic eminences GAD 67 kDa, 1:2000 (Chemicon); rabbit anti-CaBP 28 kDa, 1:3000 (Smart and Sturrock, 1979). The LGE will give rise to the (Swant); goat anti-ChAT, 1:100 (Chemicon). Sections were washed in striatum and the MGE will form the pallidum (Smart and PBS-0.1% Tween-20 and incubated with horseradish peroxidase Sturrock, 1979; Deacon et al., 1994), two structures that are part (HRP)-conjugated secondary antibodies for 2 hours at RT: HRP-goat of the basal ganglia and are involved in motion control. In both anti-mouse antibodies, 1:100 (Sigma); HRP-goat anti-rabbit antibodies, the cortex and the ganglionic eminences, two distinct proliferative 1:200 (Sigma), biotin-conjugated donkey anti-goat antibodies, 1:100 zones contribute to the generation of neuronal precursors: a (Sigma). For biotin-conjugated antibodies, the sections were incubated ventricular zone (VZ) organised as a pseudostratified with streptavidin biotinylated-HRP complex, 1:100 (Amersham) for 1 neuroepithelium and a VZ-derived subventricular zone (SVZ), hour at RT. HRP activity was detected with diaminobenzidine (DAB) or with DAB and nickel ammonium sulphate. NADPH histochemistry where dividing cells are scattered and do not form an epithelial and MAP2 immunohistochemistry were performed as described structure (Smart, 1976; Halliday and Cepko, 1992; Bhide, 1996). (Vincent et al., 1983; Anderson et al., 1997a). After their exit of the cell cycle in the VZ or in the SVZ, cells migrate laterally into the mantle where they undergo terminal In situ hybridisation differentiation. Two histochemical compartments designated In situ hybridisation was performed on 70-80 µm vibratome sections as patch (striosome) and matrix can be distinguished in the postnatal described previously (Garel et al., 1997) with the following probes: striatum. The patch and matrix compartments differ in their Ebf1, Ebf2 and Ebf3 (Garel et al., 1997); Dlx-1 (Smith-Fernandez et al., expression of neurotransmitters, neuropeptides and receptors as 1998); Dll1 (Bettenhausen et al., 1995); Dll3 (Dunwoodie et al., 1997); α β γ well as in their input from the cortex and output to the pallidum RAR , RAR , and CRABP I (Ruberte et al., 1993); RXR (Dollé et al., and substantia nigra (reviewed in Gerfen, 1992). In addition, the 1994); EphA4 (Gilardi-Hebenstreit et al., 1992); SCIP/Oct-6 (Monuki et al., 1990); Dlx-5 and Dlx-6 (Anderson et al., 1997a); cadherin-8 peak of generation of patch neurons occurs earlier than that of (Korematsu and Redies, 1997a); netrin-1 (Serafini et al., 1996). the matrix neurons (van der Kooy and Fishell, 1987). In this paper, we establish that Ebf1 plays an essential role Bromodeoxyuridine labelling and codetection in the development of the striatum. Ebf1 targeted inactivation Pregnant mice were injected intraperitoneally with 40 mg/kg of specifically affects cell differentiation at the SVZ/mantle bromodeoxyuridine (BrdU, Sigma) in PBS and killed 30 minutes or transition in the embryonic LGE, resulting in the expression of several days later. Embryo heads were fixed for 2 hours in Carnoy’s an aberrant combination of regulatory genes in the mantle. fixative, embedded in paraffin and cut in 10 µm sections. Anti-BrdU These early differentiation defects may be responsible for an immunohistochemistry was performed as described (Garel et al., 1997). increase in cell death and a dramatic atrophy observed in the For quantification of BrdU-positive cells, the scoring was performed on perinatal striatum. Interestingly, we show that defective striatal two wild-type and two homozygous mutant embryos. For each embryo, four transverse sections defined as follows were analysed: (1) the first cell differentiation is also accompanied by fasciculation and section containing the corpus callosum; (2) the first section containing navigation defects in fibre tracts passing through the striatum. the posterior branch of the anterior commissure; (3) two equidistant sections in between the previous ones. The scoring area was delimited MATERIALS AND METHODS by tangental lines drawn at the sulci between the cortex and the LGE, and between the LGE and the MGE. Scoring was performed using a Mouse line and genotyping computer-connected CCD camera placed on a Leitz DM IL microscope. Ebf1 heterozygous mice (Lin and Grosschedl, 1995) were maintained For combined BrdU labelling and in situ hybridisation, the dissected in C57/Bl6 background and crossed to produce homozygous embryos brains were fixed in 4% PFA, embedded in gelatine/albumin and cut on µ or mice. PCR genotyping of embryo yolk sac or mouse tail DNAs was a vibratome in 250 m sections. In situ hybridisation was performed on performed with the following combination of three oligonucleotides: thick floating sections, followed by BrdU immunodetection as a common 5′ oligonucleotide in the Ebf1 coding sequence (5′ described above. Thick sections were subsequently embedded in µ GCTCACTTTGAGAAGCAGCCG 3′) and two 3′ oligonucleotides paraffin and cut in 10 m sections. located in the Ebf1 5′ flanking region (5′ GGAGCCTCACCATTG- TUNEL analysis CTGTAGAG 3′) and in the neoR-coding sequence (5′ ATGGCGAT- µ GCCTGCTTGCCGAATA 3′), respectively. 35 cycles of amplification TUNEL labelling was performed on 10 m paraffin sections using the were carried out at 94¡C for 1 minute, 65¡C for 2 minutes and 72¡C In Situ Cell Death Detection Kit AP (Boehringer Mannheim). At for 2 minutes. These oligonucleotides give rise to PCR products of E15.5, TUNEL-labelled cells were scored on all the sections from different sizes with the wild-type (800 bp) and mutant (1 kb) alleles of three wild-type and three homozygous brains. At E18.5, quantification the Ebf1 gene. For staging of embryos, midday of the vaginal plug was of TUNEL-positive cells in the striatum and septum was performed considered as embryonic day 0.5 (E0.5). in every fourth section of the forebrain in four homozygous and four wild-type mice. The morphological landmarks used for delimitation Histology and immunohistochemistry of the scoring area are described in the legend of Table 1. For histological analysis, heads or dissected brains were fixed by immersion in Carnoy’s fixative (60% ethanol, 30% chloroform, 10% Fibre tracts tracing glacial acetic acid) for 2 hours, embedded in paraffin and cut in 8-10 Dissected brains were fixed by immersion or perfused with 4% µm sections. Sections were counterstained with Cresyl Violet or PFA and post-fixed overnight at 4¡C. Single crystals of the Toluidine Blue. For immunohistochemistry, embryos and mice were fluorescent carbocyanide dye DiI (1,1′-dioctadecyl 3,3,3′,3′- Ebf1 involvement in striatum development 5287 tetramethylindocarbocyanine perchlorate; Molecular Probes) were placed in the most lateral dorsal thalamus (Godement et al., 1987). After 3 weeks at 37¡C in 4% PFA to allow dye diffusion, the samples were embedded in 3% agarose, cut in 80 µm sections on a vibratome and analysed under a conventional fluorescent microscope (Leitz DM RBE).

RESULTS Ebf1 expression in the ganglionic eminences We have performed a detailed analysis of the distribution of Ebf mRNAs in the ganglionic eminences by in situ hybridisation, which confirmed the absence of Ebf2 and Ebf3 transcripts (Garel et al., 1997). Ebf1 is expressed between E11 and E17.5 in both the LGE and the MGE. In the LGE, Ebf1 mRNA was detected throughout the entire mantle region as well as in a few postmitotic cells of the SVZ that most probably correspond to neurons migrating out of the proliferative zones (Fig. 1A and Garel et al., 1997). In the MGE, Ebf1 transcripts were observed in the mantle zone immediately below the SVZ (Fig. 1B). This expression was downregulated around E17.5 (data not shown). In contrast, Ebf1 expression was maintained Fig. 1. Ebf1 expression in the ganglionic eminences. (A) Transverse in the LGE-derived striatum and at birth the entire mantle zone brain section from an E15.5 embryo hybridised with an Ebf1 probe was positive, with the exception of groups of cells located and showing the presence of Ebf1 transcripts in a few cells of the SVZ ventrolaterally and exhibiting the morphology of the forming and in the mantle of the LGE. Dotted lines indicate the VZ/SVZ or patches (Fig. 1C). Double labelling experiments with tyrosine SVZ/mantle boundaries, which can be visualised under Nomarski hydroxylase (TH) immunohistochemistry, which preferentially optics. (B) In a more caudal section, Ebf1 mRNA is detected near the labels the patch compartment at birth (van der Kooy, 1984), SVZ/mantle boundary in the MGE. (C) Coronal brain section from a confirmed that Ebf1 expression was only detected in the matrix newborn animal hybridised with the Ebf1 probe showing expression in most of the mantle of the striatum. (D) Combined Ebf1 in situ (Fig. 1D). During the first 2 postnatal weeks, Ebf1 expression hybridisation and tyrosine hydroxylase (TH) immunostaining decreased in the matrix and, in the adult, Ebf1 was not observed at high magnification indicates that the discrete groups of expressed in the striatum anymore (data not shown). cells negative for Ebf1 expression correspond to patches strongly In conclusion, Ebf1 is expressed during the entire embryonic immunoreactive for TH. DT, dorsal thalamus; LGE, lateral ganglionic development in differentiating SVZ and mantle cells in the eminence; M, mantle; MGE, medial ganglionic eminence; P, patch; LGE, suggesting that it marks a specific window during the SVZ, subventricular zone; VZ, ventricular zone. Scale bars, 200 µm. differentiation process of postmitotic striatal neurons. subventricular regions and a more diffuse staining in the LGE expansion and cell division are not affected by differentiating field of the mantle (Fig. 2A,B and data not shown). the Ebf1 mutation between E12.5 and E17.5 Using maternal injections of bromodeoxyuridine (BrdU) between To investigate the role of Ebf1 in the development of the LGE, E12.5 and E17.5, we observed that the distribution of proliferative we first examined its general morphology and the organisation cells in both the VZ and SVZ was not modified in homozygous into VZ, SVZ and mantle in Ebf1−/− embryos between E12.5 and mutant embryos (Fig. 2C,D and data not shown). Analysis of the E17.5. On Nissl-stained brain sections, the size and general number of BrdU-labelled cells in transverse sections of E15.5 appearence of the LGE in mutant homozygous embryos was LGE revealed no significant difference between wild-type and normal with a dense labelling in the ventricular and mutant embryos (wild type: 3520±635 BrdU-positive cells per section, n=8; mutant: 3492±674, n=8). Table 1. Increased cell death in the late embryonic In conclusion, these data indicate that Ebf1 is not required for striatum the formation of the LGE and the control of cell proliferation in Striatum Septum the two distinct proliferative zones. This is consistent with the −/− 814±86 (4) 270±66 (4) restriction of Ebf1 expression to postmitotic cells. Controls 441±62 (4) 235±49 (4) Ratio 1.8±9.2 1.1±0.2 Ebf1 inactivation affects the SVZ/mantle transition in the LGE TUNEL-positive cells were scored in four E18.5 homozygous mutant The expression pattern of Ebf1 suggested that its function embryos and four control embryos (two wild-type and two heterozygous embryos) on one section out of four. could be required for striatal neuron differentiation (Fig. 1). We In the striatum, apoptotic cells were counted in the area delimited by the therefore performed a systematic analysis of the expression following morphological landmarks: anteriorly, the corpus callosum, from E12.5 to E16.5 of genes expressed in restricted areas posteriorly, the posterior branch of the anterior commissure, laterally, the of the VZ, the SVZ or the mantle. These genes, whose external capsule, and ventrally, a plane parallel to the floor of the telencephalon and containing the anterior branch of the anterior commissure. transcription marks different steps in the development of On the same sections, TUNEL-positive cells were also scored in the striatal neurons, encode transcription factors (Dlx-1, Dlx-5, septum using the same ventral limit and a dorsal delimitation formed by the Dlx-6, SCIP/Oct-6), signalling molecules such as retinoid corpus callosum and the hippocampal commissure. receptors (RAR, RXR) and cellular retinoic acid binding 5288 S. Garel and others

Consequently, an aberrant combination of regulatory genes is expressed in the LGE mantle of Ebf1−/− embryos. The persistence of SVZ markers in the mantle could be explained in two ways: (1) there is a defect in striatal mantle cell differentiation; (2) striatal cells turn on markers normally with respect to their birthdate, but their migration out of the proliferative zone is accelerated and hence cells presenting SVZ characteristics are observed in the mantle. In order to discriminate between these two possibilities, we labelled cells born at E12.5, E13.5 or E14.5 using BrdU injections and examined their location in the SVZ and in the mantle at E15.5. We did not observe any difference between wild-type and mutant

Fig. 2. LGE morphology and cell division pattern in the VZ and SVZ are not affected in Ebf1−/− embryos. Adjacent transverse brain sections from wild-type (A,C) and Ebf1−/− (B,D) E15.5 embryos were processed for Nissl staining (A,B) and BrdU immunohistochemistry (C,D). Sections from wild-type and homozygous mutant embryos are presented in reversed orientation. LGE, lateral ganglionic eminence; M, mantle; MGE, medial ganglionic eminence; SVZ, subventricular zone; VZ; ventricular zone. Scale bars, 200 µm.

(CRABP I), and guidance and adhesion molecules (netrin-1, EphA4, cadherin-8). We first studied genes that are expressed in subpopulations of the VZ and in recently born cells of the SVZ, Dlx-1 and Delta like 1 (Dll1), or specifically in young postmitotic cells of the SVZ, Dll3 (Bulfone et al., 1993; Bettenhausen et al., 1995; Dunwoodie et al., 1997; Liu et al., 1997; Casarosa et al., 1999; Fig. 3A). Their patterns of expression were not modified in Ebf1−/− embryos (Fig. 3A,B and data not shown). A second series of experiments involved several genes expressed more widely in the SVZ, SCIP/Oct-6, RARα and EphA4 (Alvarez- Bolado et al., 1995; Ruberte et al., 1993; Fig. 3G), and a gene, Dlx-5, expressed at a high level in the SVZ and at a lower level in the mantle (Liu et al., 1997). From E13.5 the expression patterns of the four genes were modified in a similar manner at all rostro-caudal levels: the domain of high expression extended into a large part of the mantle (Fig. 3C-J). Finally, Fig. 3. Alterations of the pattern of gene expression in the LGE we studied markers expressed in the mantle, Dlx-6, netrin-1, mantle from Ebf1−/− embryos. Transverse brain hemisections from RARβ, RXRγ, CRABP I and cadherin-8 (Ruberte et al., 1993; wild-type (A,C,E,G,I,K,M,O) and homozygous (reverse orientation) Dollé et al., 1994; Serafini et al., 1996; Korematsu and Redies, (B,D,F,H,J,L,N,P) E14.5/15.5 embryos hybridised with Dlx-1 (A,B), 1997b; Liu et al., 1997). While Dlx-6, netrin-1, RARβ and SCIP/Oct-6 (C,D), RARα (E,F), EphA4 (G,H), Dlx-5 (I,J), Dlx-6 RXRγ mantle expression patterns were not affected by the Ebf1 (K,L), CRABP I (M,N) and cadherin-8 (O,P) probes. SCIP/Oct-6, α, mutation, the level and the domain of expression of CRABP I RAR EphA4 and Dlx-5 expression domains are expanded in the and cadherin-8 were dramatically reduced in Ebf1−/− embryos LGE mantle of the homozygous mutant (D,F,H,J, arrowheads), whereas the EphA4 expression pattern is unaffected in the MGE. (Fig. 3K-P and data not shown). Together, these data indicate Dlx-6 mantle expression is unaffected, whereas the CRABP I and that after E13.5, the transition from the SVZ to the mantle is cadherin-8 expression domains are eliminated or severely reduced dramatically affected in homozygous mutant embryos, with the (N,P, arrows). LGE, lateral ganglionic eminence; M, mantle; mcad8, persistence of SVZ-specific markers in the mantle and the cadherin-8; MGE, medial ganglionic eminence; SVZ, subventricular absence or reduced activation of at least two mantle markers. zone; VZ, ventricular zone. Scale bars, 500 µm. Ebf1 involvement in striatum development 5289

development of the mature brain, we performed a morphological analysis of the striatum in Ebf1−/− E18.5 embryos (n=8, for each type), newborns (n=11, for each type) and 3-week old mice (n=5, for each type; this latter number was limited by the low survival of the homozygous mutants). A common feature was the reduction in size of the striatum. This reduction was still limited at E18.5, but very clear at P0 (Fig. 5A,B). At P20, the striatum appeared dramatically reduced in size and presented an abnormal shape (Fig. 5K- N). The cells did not appear more densely packed in the homozygous mutant than in the wild type, suggesting that the striatum reduction reflected a decrease in the number of cells. We investigated whether this size reduction might be due to an increase in cell death. In wild-type mice, it has been estimated that 30% of the striatal neurons die during the first postnatal week, while little cell death has been observed in the embryonic striatum (Fentress et al., 1981; Fishell and van der Kooy, 1991). Using the TUNEL procedure we have compared the occurrence of apoptosis in wild-type and homozygous mutant LGE at E15.5 and E18.5. At E15.5, the entire striatum was analysed on serial sections corresponding to three embryos of each type. Only very few TUNEL-positive cells were detected, with no significant difference between control and homozygous mutant embryos (data not shown). In contrast, at E18.5 the number of apoptotic cells had notably increased in both wild-type and homozygous mutant embryos (Fig. 6). In addition, a significant difference was observed at this stage between Ebf1−/− and control striatum: the number of apoptotic cells was higher in homozygous mutants by a factor of 1.8±0.2- fold (Table 1; n=4 for each type). Furthermore, large number of Fig. 4. Normal migration to the mantle but altered differentiation of the apoptotic cells in the homozygous mutant embryos were LGE cells. Transverse brain sections from wild-type (A,C,E) and located in the lateral (Fig. 6B) and ventral (Fig. 6D) areas of homozygous mutant (B,D,F) E15.5 embryos whose mothers were the striatum, whereas very few TUNEL-positive cells were injected with BrdU at E13.5. The sections were either processed for detected in these regions in wild-type embryos (Fig. 6A,C). BrdU immunohistochemistry alone (A,B), or for combined BrdU Finally, in the septum, the number of apoptotic cells was similar immunohistochemistry and in situ hybridisation (C-F). (A,B) BrdU- in Ebf1−/− and wild-type embryos (Fig. 6A,B and Table 1), positive cells (black dots), corresponding to cells born at E13.5, show a similar distribution in the SVZ and in the mantle of wild-type and pointing out that increased cell death in the basal telencephalon Ebf1−/− embryos. (C-F) In the LGE mantle of the homozygous of homozygous mutant embryos was specific to the striatum. In conclusion, this analysis shows a specific increase in cell death mutant, some of the BrdU-positive cells (brown dots) are colabelled − − by the SCIP/Oct-6 probe (arrowheads in D) or by the Dlx-5 probe in the striatum of late Ebf1 / embryos, providing a likely (arrowheads in F), whereas wild-type mantle cells express only very explanation for the severe reduction in size observed after birth. low levels of these genes (C,E). M, mantle; SVZ, subventricular zone. Scale bars, 500 µm. Cell type specification and compartmentation in the mutant striatum embryos, indicating that the time course of cell migration from To determine whether Ebf1 inactivation affects specific cell the proliferative zones towards the mantle was not affected in populations in the LGE-derived striatum, we analysed the Ebf1−/− embryos (Fig. 4A,B and data not shown). In addition, distribution of subsets of striatal cells: γ-aminobutyric acidic- using combined BrdU labelling and in situ hybridisation with positive (GABAergic) neurons, which constitute the large the SCIP/Oct-6 or Dlx-5 probes, we observed that, at E15.5, majority of striatal neurons (Kita and Kitaï, 1988), cholinergic some cells born at E13.5 and located in the mantle maintained and somatostatin-containing interneurons (Vincent et al., 1983) SCIP/Oct-6 or Dlx-5 expression in the homozygous mutant and astrocytes (Anderson et al., 1997a). These populations were LGE, whereas these cells had downregulated these markers in present in the striatum of Ebf1−/− newborns and young mice and wild-type embryos (Fig. 4C-F). In conclusion, these data exhibited a normal density (Fig. 5Q-T and data not shown). indicate that Ebf1 mutation specifically affects the differentiation We next examined the patch/matrix compartmentation. At process, which is coincidental with the migration from the SVZ birth, using TH or DARPP-32 immunohistochemistry (van der towards the mantle, leading to an abnormal expression pattern Kooy, 1984; Foster et al., 1987), we observed that the patches of genes in the mantle. were present in the striatum of Ebf1−/− newborns, although their distribution appeared slightly modified (Fig. 5C,D and Striatum atrophy and increased cell death in data not shown). In absence of another known matrix marker perinatal mice at birth, we monitored the transcriptional activity of the To investigate the effects of the early LGE phenotype on the Ebf1 gene, which we have shown to be restricted to this 5290 S. Garel and others compartment (Fig. 1C,D). We made use of the fact that neo other morphological alteration of the striatum was detected antisense expression faithfully recapitulates the normal Ebf1 (Fig. 7A,B). Finally, the posterior branch of the anterior expression pattern in heterozygous animals (data not shown). In homozygous mutant newborns, the neo probe labelled a much smaller domain than in heterozygous mice (Fig. 5E,F), suggesting that the matrix compartment is severely reduced at birth. At P20, the use of calbindin (CaBP) immunoreactivity, which labels the matrix at that age (Liu and Graybiel, 1992), confirmed a strong reduction of this compartment in Ebf1−/− mice (data not shown). Taken together, these data indicate that none of the studied cell populations is eliminated in Ebf1 mutant mice and the only imbalance that we have noticed is a dramatic reduction of the matrix, whereas the patch compartment is relatively preserved. Finally, as it has been shown that the ganglionic eminences are also the site of production of interneurons that populate the cortex and the olfactory bulb (De Carlos et al., 1996; Anderson et al., 1997b; Tamamaki et al., 1997; reviewed in Alvarez-Buylla, 1997, Sussel et al., 1999; Wichterle et al., 1999), we examined these two neuronal populations in Ebf1−/− mice using glutamic acid decarboxylase (GAD) immunohistochemistry. We found that these populations were normally present at their final locations (Fig. 5G-J), suggesting that Ebf1 inactivation does not affect tangentially migrating cells originating from or passing through the LGE. Therefore, among LGE-derived cells, Ebf1 is more specifically involved in striatal cell differentiation. Abnormal navigation and fasciculation of the internal capsule fibres As we observed modifications in the expression pattern of guidance and adhesion molecules in the LGE of Ebf1−/− embryos, we examined whether fibre tracts passing through the striatum are affected. The internal capsule is a major fibre tract that travels through the striatum and is composed of cortical axons projecting subcortically and axons originating Fig. 5. Morphological alterations in the striatum from Ebf1−/− mice. Coronal brain from various CNS structures projecting sections from newborns (A-J), P20 (K-R) and P14 (S,T) mice. In each column sections towards the cortex. Fibre bundles running in from heterozygous or wild-type mice are presented on the left and sections of the caudal part of the internal capsule homozygous mutant animals on the right in reverse orientation. (A,B) Nissl staining. fasciculate before turning medially and Note the reduction in size of the striatum. (C,D) DARPP-32 immunostaining. Strongly entering the pallidum. In Ebf1−/− animals, the positive patches are detected both in heterozygous and homozygous embryos. (E,F) In fibres of the internal capsule exhibited an situ hybridisation with a neo sense riboprobe to visualise Ebf1 transcription. The patches abnormal fasciculation: the bundles of fibres are observed as neo-negative domains. The domain of Neo sense expression is severely were generally larger than in wild-type reduced in the homozygous newborn (arrows). (G-J) GAD immunostaining showing the normal presence of GABAergic interneurons in the olfactory bulb (G,H) and outer layer animals and their distribution appeared less − − (arrows) of the cortex (I,J) in Ebf1 / newborns. (K,L) Nissl staining. (M-P) TH regular (Fig. 5A,B,M-P). In the caudal part, immunohistochemistry. Note the dramatic atrophy of the striatum and the modification of where the internal capsule normally forms a the internal capsule pattern in Ebf1−/− mice. (Q,R) NADPH diaphorase histochemistry single large bundle, a disorganisation was also labelling somatostatin-containing striatal interneurons. (S,T) ChAT observed (Fig. 7C,D). This fasciculation immunohistochemistry staining striatal cholinergic interneurons. CX, cortex; IC, internal defect was already visible at E15.5, before any capsule; NADPH dia., NADPH diaphorase; P, patch; Str, striatum. Scale bars, 500 µm. Ebf1 involvement in striatum development 5291

Fig. 6. Increased apoptosis in the Ebf1−/− striatum. Transverse brain sections from wild-type (A,C) and homozygous mutant (reverse orientation) (B,D) E18.5 embryos were processed for the TUNEL reaction. Blue dots correspond to apoptotic cells. In the wild-type embryo, very few apoptotic figures are detected in the lateral region of the striatum (A, arrowhead), whereas several of them are observed in this region of the homozygous mutant striatum (B, arrowheads). A similar observation can be made in the ventralmost part of the striatum where many TUNEL-positive cells are detected in Ebf1−/− embryos close to the posterior branch of the anterior commissure (arrowheads in C,D). acP, posterior branch of the anterior commissure; Se, septum. Scale bars, 200 µm.

differentiation of striatal cells while they migrate from the SVZ to the mantle, resulting in the expression of an abnormal combination of regulatory genes in the mantle. This early phenotype leads to major alterations in the mature ventral telencephalon: (1) an increase in cell death, which correlates commissure, which runs underneath the striatum, was also with a dramatic reduction in size of the postnatal striatum, and affected, a large majority of the fibres failing to cross the (2) axonal navigation defects in the fibres of the internal midline (data not shown). capsule passing throughout the striatum. To further characterise the phenotype affecting telencephalic fibre tracts, we traced the thalamocortical fibres, which are an Ebf1 controls the SVZ/mantle transition in the LGE important component of the internal capsule, by placing DiI We have shown that Ebf1 inactivation specifically affects a step crystals in the dorsolateral thalamus of E15.5 and E18.5 of striatal cell differentiation that is concomitant with the embryos (Fig. 7E-J). In E15.5 Ebf1−/− embryos, DiI-labelled migration from the SVZ towards the mantle. Several pieces of fibres presented a normal navigation and fasciculation data suggest that the progenitors in the VZ and SVZ are not before entering the ganglionic eminences (data not shown). In the LGE, these fibres spread out in the mantle and entered the ventralmost part (Fig. 7H), whereas in wild-type embryos these fibres were restricted to the dorsal part (Fig. 7E,G). At E18.5, the characteristic fan- shaped distribution of the thalamocortical fibres in the striatal mantle was completely disorganised in the homozygous mutant (Fig. 7I,J). However, after exiting the striatum DiI-labelled fibres refasciculated and reached the cortex (data not shown). In conclusion, these results indicate that thalamocortical fibres are specifically affected Fig. 7. Navigation and fasciculation defects in the internal capsule of homozygous mutant embryos. as they travel through the (A-D) Morphological analysis of the internal capsule in transversal brain sections from wild type (A,C) striatum primordium, and this and homozygous mutant (B,D) embryos. The sections were processed for MAP2 immunohistochemistry phenotype is detected before and the fibre tracts appear in white. (A,B) At E15.5 the internal capsule is highly disorganised in the any other morphological defect mutant, presenting additional ventral branches (B, arrowheads). (C,D) At E18.5 fasciculation of the can be observed in the LGE. internal capsule is also abnormal in the Ebf1−/− embryo. (E,F) Schematic representations of the connections between the cortex and the thalamus at E15.5 (E) and E18.5 (F). At E15.5 cortico-fugal fibres are shown in yellow and thalamocortical in blue, and at E18.5 interconnections between the cortex DISCUSSION and the dorsal thalamus are drawn in green (adapted from Métin and Godement, 1996 and Molnar et al., 1998). Arrows indicate the positions where DiI crystals were placed in (G,H) and (I,J), respectively. Dashed boxes indicate the zones pictured in (G-J). (G,H) DiI labelling of the connections between the In this paper, we have cortex and the thalamus in E15.5 wild-type (G) and Ebf1−/− (H) embryos. Instead of forming a condensed established that Ebf1 plays tract, the mutant fibres show an abnormal spreading in the mantle of the LGE. (I,J) DiI labelling at E18.5. a major role in the While the fibres form a nicely arranged fan-shaped tract in the wild-type embryo (I), they show a severe development of the striatum. disorganisation in the mutant (J). CX, cortex; GE, ganglionic eminence; IC, internal capsule; L, lateral Ebf1 inactivation affects the ganglionic eminence; M, medial ganglionic eminence; TH, thalamus. Scale bars, 500 µm. 5292 S. Garel and others affected in homozygous mutant embryos: (1) the cell division in homozygous mutant embryos (Fig. 5G-J). It will be of pattern in the VZ and SVZ is not modified; (2) the early particular interest to determine whether Ebf1 constitutes a expansion of the LGE is normal, suggesting that cell production downstream target of Dlx-1 and Dlx-2, since the combined is not affected; (3) no molecular alterations were observed in inactivation of these factors, like the Ebf1 mutation, more the VZ and SVZ. In contrast, gene expression in the mantle is severely affects the formation of the matrix compartment clearly modified: the expression of several SVZ markers is (Anderson et al., 1997a). abnormally maintained and two mantle markers, cadherin-8 and CRABP I, are not activated appropriately. Since radial cell Ebf1 inactivation affects cell survival in the migration from the SVZ towards the mantle is not affected by developing striatum the mutation, the maintenance of SVZ genes expression in the We have shown that apoptosis is augmented in the striatum mantle must reflect an abnormal differentiation process at the during late embryogenesis in homozygous mutants, suggesting SVZ/mantle transition. Some mantle markers are nevertheless that an increase in cell death is responsible for the reduction in expressed in the mantle of homozygous embryos (Fig. 3K,L and size observed after birth. Since the size reduction appears to be data not shown), indicating that many mantle cells abnormally more dramatic at P20 than at birth, it is likely that increased coexpress both SVZ and mantle markers. In conclusion, apoptosis is maintained in homozygous mutant animals during Ebf1−/− striatal mantle cells present an aberrant molecular early postnatal development. This increased apoptosis may be identity, with combined SVZ and mantle characteristics, and a direct consequence of the aberrant molecular identity of LGE our analysis has revealed a major role of Ebf1 in the regulation mantle cells observed at early stages. An alternative, but not of the SVZ/mantle transition. exclusive, possibility is suggested by the coincidence between the beginning of the phase of increased cell death and the The genetic hierarchy governing neurogenesis in establishment of connections between late-born striatal cells the ventral telencephalon and other areas of the brain, including the input from the In the wild-type striatum, the downregulation of SVZ markers neocortex and the substantia nigra and the output to the and the activation of mantle genes are concomitant. Interestingly, substantia nigra (Gerfen et al., 1987; Fishell and van der Kooy, the Ebf1 mutation uncouples these two processes, suggesting 1989; Sheth et al., 1998). As the molecular alterations in the that they are controlled by at least two partially independent LGE include modifications of the expression of adhesion and regulatory pathways. Furthermore, while Ebf1 has previously guidance molecules, the formation of synapses may be only been shown to function as a transcriptional activator, in B perturbed, possibly affecting striatal neurons survival (Fishell cells and in olfactory neurons (Hagman et al., 1993, 1995; Wang and van der Kooy, 1991 and references therein). and Reed, 1993; Wang et al., 1997), the present work points to An additional question is why the effect of the Ebf1 mutation an involvement of Ebf1 in the activation of two mantle markers appears more dramatic on the matrix than on the patch as well as the downregulation of several SVZ markers. Further compartment. The patch compartment is in large part experiments will determine whether these genes constitute direct composed of early born cells (E11-E12.5) (van der Kooy and transcriptional targets of Ebf1. Fishell, 1987). The relative preservation of the patch Among the putative Ebf1 transcription targets, those compartment is therefore consistent with the appearance of encoding transcription factors (Dlx-5, SCIP/Oct6) and proteins molecular defects only after E13.5. involved in retinoid signalling (RARα, CRABP I) are most likely to affect cell type specification in the striatum. In this Impaired differentiation in the LGE leads to fibre respect, it is very interesting to note that radial glia in the LGE tract navigation defects produce retinoids and that retinoids enhance striatal neuron We have shown that Ebf1 inactivation affects both axonal differentiation in vitro (Toresson et al., 1999). An abnormal navigation and fasciculation of thalamocortical fibres in the combination of retinoic acid receptors and the absence of striatum, indicating that interactions between these fibres and CRABP I in the Ebf1−/− LGE mantle may therefore affect the the striatum are specifically perturbed. Our results suggest that terminal differentiation of neurons migrating radially from the thalamocortical fibres navigation defects are due to the early SVZ towards the mantle and contribute to the later phenotypes. LGE differentiation phenotype. In particular, EphA4 and During the last years, inactivation of several genes, the cadherin-8 expression domains are modified in the LGE of genes Gsh-2, Dlx-1 and Dlx-2, the LIM homeobox Ebf1−/− embryos. An attractive possibility is that the persistence gene Lhx-2 and the bHLH gene Mash-1, has demonstrated a role in the LGE mantle of SVZ-specific cues such as EphA4 may be of these factors in the specification of VZ and SVZ neuronal responsible for the ventral misrouting of thalamocortical fibres, precursors in the LGE (Anderson et al., 1997a; Porter et al., 1997; which normally travel just underneath the SVZ/mantle Szucsik et al., 1997; Casarosa et al., 1999). Hence, Mash-1 boundary. Furthermore, cadherin-8, which is normally present targeting affects the VZ progenitors, and double Dlx-1 and Dlx- in both the striatum and the thalamocortical fibres, could 2 mutation impairs the migration and differentiation of SVZ mediate fasciculation through homophilic interactions progenitors. These mutations affect both radially migrating (Korematsu et al., 1998). In Ebf1−/− embryos, cadherin-8 striatal cells and SVZ-derived tangentially migrating cells that expression is not affected in the thalamus (data not shown), will populate the cortex and olfactory bulb (Anderson et al., while it is strongly reduced in the LGE mantle. This may lead 1997a,b; Casarosa et al., 1999). Factors specifically governing to a defect in the interactions between the fibres and the cell differentiation in the LGE have not been characterised environment, and in consequence favour fibre/fibre interactions, previously. Our analysis shows that Ebf1 plays precisely such a resulting in the formation of larger fibre bundles. While we key role. Furthermore, Ebf1 is specifically involved in striatal cell cannot exclude that Ebf1 inactivation directly affects the differentiation, since tangentially migrating cells are not affected differentiation of thalamic neurons and hence the navigation of Ebf1 involvement in striatum development 5293 their axons, our results do not favour this hypothesis. Ebf1 is Association pour la Recherche sur le Cancer, the Fondation pour la expressed in the dorsal thalamus only until E14.5 (Fig. 1B, Recherche Médicale and the Ligue Nationale Française Contre le Garel et al., 1997) and we did not detect any morphological Cancer. alterations nor any modifications in the expression of several adhesion and guidance molecules (EphA4, ephrin-A2, ephrin- REFERENCES A5, ephrin-B3 and cadherin-8) in the thalamus of homozygous mutant embryos (data not shown). Alvarez-Bolado, G., Rosenfeld, M. G. and Swanson, L. W. (1995). Model To date, the ganglionic eminences have only been shown to of forebrain regionalization based on spatiotemporal patterns of POU-III homeobox gene expression, birthdates, and morphological features. J. be involved in the long-range attraction of the fibres that Comp. Neurol. 355, 237-295. constitute the internal capsule (Métin et al., 1997; Richards et Alvarez-Buylla A. (1997). Mechanism of migration of olfactory bulb al., 1997; Braisted et al., 1999). The data presented in this interneurons. Sem. Cell. Dev. Biol. 8, 207-213. report suggest that the LGE also plays a major role in the local Anderson, S. A., Qiu, M., Bulfone, A., Eisenstat, D. D., Meneses, J., control of both axonal pathfinding and fasciculation, via the Pedersen, R. and Rubenstein, J. L. (1997a). Mutations of the homeobox genes Dlx-1 and Dlx-2 disrupt the striatal subventricular zone and restricted expression of specific guidance cues in the SVZ or differentiation of late born striatal neurons. Neuron 19, 27-37. in the mantle. Anderson, S. A., Eisenstat, D. D., Shi, L. and Rubenstein, J. L. (1997b). Interneuron migration from basal forebrain to neocortex: dependence on Dlx The Ebf multigene family and neuronal genes. Science 278, 474-476. differentiation Bettenhausen, B., Hrabe de Angelis, M., Simon, D., Guenet, J. L. and Gossler, A. (1995). Transient and restricted expression during mouse Ebf1, Ebf2 and Ebf3 are transiently expressed in a large number embryogenesis of Dll1, a murine gene closely related to Drosophila Delta. of postmitotic neurons of the CNS with very dynamic and Development 121, 2407-2418. overlapping expression patterns (Davis and Reed, 1996; Garel Bhide, P. G. (1996). Cell cycle kinetics in the embryonic mouse corpus et al., 1997; Malgaretti et al., 1997; Wang et al., 1997). We striatum. J. Comp. Neurol. 374, 506-522. Braisted, J. E., Tuttle, R. and O’Leary, D. D. M. (1999). Thalamocortical have shown that Ebf1 plays a major role in striatum axons are influenced by chemorepellent and chemoattractant activities differentiation, where it is the only member of the Ebf family localized to decision points along their path. Dev. Biol. 208, 430-440. to be expressed. This observation raises the possibility that Bulfone, A., Puelles, L., Porteus, M. H., Frohman, M. A., Martin, G. R. Ebf1 may be exerting a similar role in other neuronal and Rubenstein, J. L. (1993). Spatially restricted expression of Dlx-1, Dlx- 2 (Tes-1), Gbx-2, and Wnt- 3 in the embryonic day 12.5 mouse forebrain populations, but that the expression of the other family defines potential transverse and longitudinal segmental boundaries. J. members may compensate in its absence. This idea is Neurosci. 13, 3155-3172. supported by our observation that the facial motor nucleus, Casarosa, S., Fode, C. and Guillemot, F. (1999). Mash1 regulates where Ebf1 is also the only member of the family expressed, neurogenesis in the ventral telencephalon. Development 126, 525-534. develops abnormally in Ebf1−/− embryos (S. Garel et al., Cau, E., Gradwohl, G., Fode, C. and Guillemot, F. (1997). Mash1 activates a cascade of bHLH regulators in olfactory neuron progenitors. Development unpublished). Furthermore, mutations in the C. elegans Ebf 124, 1611-1621. ortholog, unc-3, affect neuronal differentiation and ventral Davis, J. A. and Reed, R. R. (1996). Role of Olf-1 and Pax-6 transcription nerve cord fasciculation (Prasad et al., 1998), a phenotype factors in neurodevelopment. J. Neurosci. 16, 5082-5094. strikingly similar to the loss of Ebf1 function in the striatum. De Carlos, J. A., Lopez-Mascaraque, L. and Valverde, F. (1996). Dynamics of cell migration from the lateral ganglionic eminence in the rat. J. Neurosci. Therefore, aspects of the roles of Ebf genes in neuronal 16, 6146-6156. differentiation have been conserved during the course of Deacon, T. W., Pakzaban, P. and Isacson, O. (1994). The lateral ganglionic evolution. eminence is the origin of cells committed to striatal phenotypes: neural In the mouse CNS, Ebf2 is transiently expressed in young transplantation and developmental evidence. Brain Res. 668, 211-219. postmitotic cells whereas Ebf1 and Ebf3 are expressed in more Dollé, P., Fraulob, V., Kastner, P. and Chambon, P. (1994). Developmental expression of murine (RXR) genes. Mech. Dev. 45, 91- differentiated cells (Garel et al., 1997). These expression patterns 104. suggest that Ebf2 may be involved in the control of earlier steps Dubois, L., Bally-Cuif, L., Crozatier, M., Moreau, J., Paquereau, L. and of differentiation than Ebf1 and Ebf3. This idea is consistent with Vincent, A. (1998). XCoe2, a transcription factor of the Col/Olf-1/EBF the modifications observed in early steps of primary neuron family involved in the specification of primary neurons in Xenopus. Curr. Biol. 8, 199-209. differentiation in Xenopus embryos following alterations of Dunwoodie, S. L., Henrique, D., Harrison, S. M. and Beddington, R. S. XCoe2 (Ebf2 ortholog) activity (Dubois et al., 1998). The (1997). Mouse Dll3: a novel divergent Delta gene which may complement generation of null mutations in Ebf2 and Ebf3 will allow the the function of other Delta homologues during early pattern formation in precise characterisation of the function of each Ebf gene and the mouse embryo. Development 124, 3065-76. combination of these mutations may provide a demonstration of Fentress, J. C., Stanfield, B. B. and Cowan, W. M. (1981). Observation on the development of the striatum in mice and rats. Anat. Embryol. 163, 275-298. a general role of this family in neuronal differentiation. Fishell, G. and van der Kooy, D. (1989). Pattern formation in the striatum: developmental changes in the distribution of striatonigral projections. Brain − − We wish to thank H. Lin and M. O’Riordan for providing Ebf1 / Res. Dev. Brain Res. 45, 239-255. embryos and mice, C. Métin, J. L. Rubenstein and M. Wassef for Fishell, G. and van der Kooy, D. (1991). Pattern formation in the striatum: helpful discussions and thoughtful comments on the manuscript, and neurons with early projections to the substantia nigra survive the cell death I. Fariñas for communication of unpublished data. We would also like period. J. Comp. Neurol. 312, 33-42. to acknowledge the following people for their gifts of probes or Fode, C., Gradwohl, G., Morin, X., Dierich, A., LeMeur, M., Goridis, C. antibodies: R. Beddington, P. Chambon, S. Dunwoodie, J. A. Girault, and Guillemot, F. (1998). The bHLH protein NEUROGENIN 2 is a determination factor for epibranchial placode-derived sensory neurons. H. Gronemeyer, D. Henrique, G. Lemke, C. Redies, J. Rubenstein, A. Neuron 20, 483-494. Smith-Fernandez, M. Tessier-Lavigne and M. Wassef. This work was Foster, G. A., Schultzberg, M., Hokfelt, T., Goldstein, M., Hemmings, H. supported by grants from the Institut National de la Santé et de la C. J., Ouimet, C. C., Walaas, S. I. and Greengard, P. (1987). Recherche Médicale, the Ministère de l’Education Nationale, de la Development of a dopamine-and cyclic adenosine 3′:5′-monophosphate- Recherche et de la Technologie, the European Community, the regulated phosphoprotein (DARPP-32) in the prenatal central nervous 5294 S. Garel and others

system, and its relationship to the arrival of presumptive dopaminergic Miller, B., Chou, L. and Finlay, B. L. (1993). The early development of innervation. J. Neurosci. 7, 1994-2018. thalamocortical and corticothalamic projections. J. Comp. Neurol. 335, 16- Garel, S., Marin, F., Mattei, M. G., Vesque, C., Vincent, A. and Charnay, 41. P. (1997). Family of Ebf/Olf-1-related genes potentially involved in Molnar, Z., Adams, R. and Blakemore, C. (1998). Mechanisms underlying neuronal differentiation and regional specification in the central nervous the early establishment of thalamocortical connections in the rat. J. system. Dev. Dyn. 210, 191-205. Neurosci. 18, 5723-5745. Gerfen, C. R., Baimbridge, K. G. and Thibault, J. (1987). The neostriatal Monuki, E. S., Kuhn, R., Weinmaster, G., Trapp, B. D. and Lemke, G. mosaic: III. Biochemical and developmental dissociation of patch-matrix (1990). Expression and activity of the POU transcription factor SCIP. mesostriatal systems. J. Neurosci. 7, 3935-3944. Science 249, 1300-1303. Gerfen, C. R. (1992). The neostriatal mosaic: multiple levels of Porter, F. D., Drago, J., Xu, Y., Cheema, S. S., Wassif, C., Huang, S. P., compartmental organization in the basal ganglia. Annu. Rev. Neurosci. 15, Lee, E., Grinberg, A., Massalas, J. S., Bodine, D., Alt, F. and Westphal, 285-320. H. (1997). Lhx2, a LIM homeobox gene, is required for eye, forebrain, and Gilardi-Hebenstreit, P., Nieto, M. A., Frain, M., Mattei, M. G., Chestier, definitive erythrocyte development. Development 124, 2935-2944. A., Wilkinson, D. G. and Charnay, P. (1992). An Eph-related receptor Prasad, B. C., Ye, B., Zackhary, R., Schrader, K., Seydoux, G. and Reed, protein tyrosine kinase gene segmentally expressed in the developing mouse R. R. (1998). unc-3, a gene required for axonal guidance in Caenorhabditis hindbrain. Oncogene 7, 2499-2506. elegans, encodes a member of the O/E family of transcription factors. Godement, P., Vanselow, J., Thanos, S. and Bonhoeffer, F. (1987). A study Development 125, 1561-1568. in developing visual systems with a new method of staining neurones and Richards, L. J., Koester, S. E., Tuttle, R. and O’Leary, D. D. (1997). their processes in fixed tissue. Development 101, 697-713. Directed growth of early cortical axons is influenced by a chemoattractant Guillemot, F., Lo, L. C., Johnson, J. E., Auerbach, A., Anderson, D. J. and released from an intermediate target. J. Neurosci. 17, 2445-2458. Joyner, A. L. (1993). Mammalian achaete-scute homolog 1 is required for Ruberte, E., Friederich, V., Chambon, P. and Morriss-Kay, G. (1993). the early development of olfactory and autonomic neurons. Cell 75, 463-476. Retinoic acid receptors and cellular retinoid binding proteins. III. Their Hagman, J., Travis, A. and Grosschedl R. (1991). A novel lineage-specific differential transcript distribution during mouse nervous system nuclear factor regulates mb-1 gene transcription at the early stages of B cell development. Development 118, 267-282. differentiation. EMBO J. 10, 3409-3417. Serafini, T., Colamarino, S. A., Leonardo, E. D., Wang, H., Beddington, Hagman, J., Belanger, C., Travis, A., Turck, C. W. and Grosschedl, R. R., Skarnes, W. C. and Tessier-Lavigne, M. (1996). Netrin-1 is required (1993). Cloning and functional characterization of early B-cell factor, a for commissural axon guidance in the developing vertebrate nervous system. regulator of lymphocyte-specific gene expression. Genes Dev. 7, 760-773. Cell 87, 1001-1014. Hagman, J., Gutch, M. J., Lin, H. and Grosschedl, R. (1995). EBF contains Sheth, A. N., McKee, M. L. and Bhide, P. G. (1998). The sequence of a novel zinc coordination motif and multiple dimerization and formation and development of corticostriate connections in mice. Dev. transcriptional activation domains. EMBO J. 14, 2907-2916. Neurosci. 20, 98-112. Halliday, A. L. and Cepko, C. L. (1992). Generation and migration of cells Smart, I. H. M. (1976). A pilot study of cell production by the ganglionic in the developing striatum. Neuron 9, 15-26. eminences of the developing mouse brain. J. Anat. 121, 71-84. Hirsch, M. R., Tiveron, M. C., Guillemot, F., Brunet, J. F. and Goridis, C. Smart, I. H. M. and Sturrock, R. R. (1979). Ontogeny of the neostriatum. (1998). Control of noradrenergic differentiation and Phox2a expression by In The Neostriatum (ed. I. Divak and R. G. E. Oberg,), pp. 127-146. MASH1 in the central and peripheral nervous system. Development 125, Pergamon, New York. 599-608. Smith-Fernandez, A., Pieau, C., Reperant, J., Boncinelli, E. and Wassef, Kita, H. and Kitaï, S. T. (1988). Glutamate decarboxylase immunoreactive M. (1998). Expression of the Emx-1 and Dlx-1 homeobox genes define three neurons in rat neostriatum: their morphological types and populations. Brain molecularly distinct domains in the telencephalon of mouse, chick, turtle Res. 447, 346-352. and frog embryos: implications for the evolution of telencephalic Korematsu, K. and Redies, C. (1997a). Restricted expression of cadherin-8 subdivisions in amniotes. Development 125, 2099-2111. in segmental and functional subdivisions of the embryonic mouse brain. Sussel, L., Marin, O., Kimura, S. and Rubenstein, J. L. (1999). Loss of Dev. Dyn. 208, 178-189. Nkx2.1 homeobox gene function results in a ventral to dorsal molecular Korematsu, K. and Redies, C. (1997b). Expression of cadherin-8 mRNA in respecification within the basal telencephalon: evidence for a transformation the developing mouse central nervous system. J. Comp. Neurol. 387, 291- of the pallidum into the striatum. Development 126, 3359-3370. 306. Szucsik, J. C., Witte, D. P., Li, H., Pixley, S. K., Small, K. M. and Potter, Korematsu, K., Nishi, T., Okamura, A., Goto, S., Morioka, M., Hamada, S. S. (1997). Altered forebrain and hindbrain development in mice mutant J. and Ushio, Y. (1998). Cadherin-8 protein expression in gray matter for the Gsh-2 homeobox gene. Dev. Biol. 191, 230-242. structures and nerve fibres of the neonatal and adult mouse brain. Tamamaki, N., Fujimori, K. E. and Takauji, R. (1997). Origin and route of Neuroscience 87, 303-315. tangentially migrating neurons in the developing neocortical intermediate Lin, H. and Grosschedl, R. (1995). Failure of B-cell differentiation in mice zone. J. Neurosci. 17, 8313-8323. lacking the transcription factor EBF. Nature 376, 263-267. Toresson, H., Mata de Urquiza, A., Fagerström, C., Perlmann, T. and Liu, F. C. and Graybiel, A. M. (1992). Heterogeneous development of Campbell, K. (1999). Retinoids are produced by glia in the lateral calbindin-D28K expression in the striatal matrix. J. Comp. Neurol. 320, 304- ganglionic eminence and regulate striatal neuron differentiation. 322. Development 126, 1317-1326. Liu, J. K., Ghattas, I., Liu, S., Chen, S. and Rubenstein, J. L. (1997). Dlx van der Kooy, D. (1984). Developmental relationships between opiate genes encode DNA-binding proteins that are expressed in an overlapping receptors and dopamine in the formation of caudate-putamen patches. Brain and sequential pattern during basal ganglia differentiation. Dev. Dyn. 210, Res. 316, 300-303. 498-512. van der Kooy, D. and Fishell, G. (1987). Neuronal birthdate underlies the Ma, Q., Chen, Z., del Barco Barrantes, I., de la Pompa, J. L. and development of striatal compartments. Brain Res. 401, 155-61. Anderson, D. J. (1998). neurogenin1 is essential for the determination of Vincent, S. R., Johansson, O., Hokfelt, T., Skirboll, L., Elde, R. P., neuronal precursors for proximal cranial sensory ganglia. Neuron 20, 469- Terenius, L., Kimmel, J. and Goldstein, M. (1983). NADPH-diaphorase: 482. a selective histochemical marker for striatal neurons containing both Malgaretti, N., Pozzoli, O., Bosetti, A., Corradi, A., Ciarmatori, S., somatostatin- and avian pancreatic polypeptide (APP)- like Panigada, M., Bianchi, M. E., Martinez, S. and Consalez, G. G. (1997). immunoreactivities. J. Comp. Neurol. 217, 252-263. Mmot1, a new helix-loop-helix transcription factor gene displaying a sharp Wang, M. M. and Reed, R. R. (1993). Molecular cloning of the olfactory expression boundary in the embryonic mouse brain. J. Biol. Chem. 272, neuronal transcription factor Olf-1 by genetic selection in yeast. Nature 364, 17632-17639. 121-126. Métin, C., Deleglise, D., Serafini, T., Kennedy, T. E. and Tessier-Lavigne, Wang, S. S., Tsai, R. Y. and Reed, R. R. (1997). The characterization of the M. (1997). A role for netrin-1 in the guidance of cortical efferents. Olf-1/EBF-like HLH transcription factor family: implications in olfactory Development 124, 5063-5074. gene regulation and neuronal development. J. Neurosci. 17, 4149-4158. Métin, C. and Godement, P. (1996). The ganglionic eminence may be an Wichterle, H., Garcia-Verdugo, J. M., Herrera, D.G. and Alvarez-Buylla, intermediate target for corticofugal and thalamocortical axons. J. Neurosci. A. (1999). Young neurons from medial ganglionic eminence disperse in 16, 3219-3235. adult and embryonic brain. Nature Neurosci. 2, 461-466.