Abstract

In this study we examine possible origins and migratory routes of human cortical neurons, with special emphasis on the preplate and layer I. In embryonic stages, two main cell types, Cajal–Retzius cells, and cells labeled with interneuron markers (calretinin, calbindin and GABA), were present in the preplate layer. In addition, a number of preplate GABAergic cells co-expressed either Nkx2.1 or Dlx transcription factors, findings consistent with their origin in the ganglionic eminence and subsequent tangential migration to the layer I. The orientation of the leading process indicates that some of these cells descend to the cortical plate. However, the finding of radially oriented GABAergic, NKX2.1+ and DLX+ cells in the cortical ventricular zone, argues that, unlike in rodents, a significant subpopulation of these cells originates in the cortical ventricular zone. In embryonic stages, expression of Reelin in Cajal–Retzius cells as well as Reelin/DLX2+ cells in the embryonic ganglionic eminence and the olfactory region, suggest that these cells in human may have diverse origins. In later fetal stages in human (17–22 gestational weeks) layer I and the newly formed subpial granular layer, contained a population of small interneurons that originated mainly in the lateral ganglionic eminence, since the majority of these cells were double-labeled with DLX/GABA, and rarely with NKX2.1/GABA. Therefore, neurons in the human cortical layer I are heterogeneous, with more complex origin and migratory routes than in rodents. In addition to the ganglionic eminence, both the expended subventricular zone and subpial granular layer, contribute to the neuronal population of the developing layer I and underlining cortical plate.

Introduction

Because of the recent advances made in understanding the development of cortical interneurons in rodents (Parnavelas, 2000; Marin and Rubenstein, 2001), non-human primates (Zecevic and Rakic, 2001) and humans (Letinic et al., 2002), it is important to re-examine the development of human layer I. There is good evidence that layer I in primates contains a greater variety of neurons, including interneurons, in comparison to subprimate species (Zecevic and Milosevic, 1997; Meyer and Goffinet, 1998; Zecevic et al., 1999; Zecevic and Rakic, 2001). The available literature indicates several possible sources of cortical cells, some of which require migration of precursor cells through divisional boundaries. One such source is the ganglionic eminence (GE) of the ventral telencephalon, which has two parts, medial (MGE) and lateral (LGE), that are distinguishable in humans from Carnegie stage 18 (O’Rahilly and Muller, 1994; Grasbon-Frodl et al., 1996). In primates, the GE was considered to be the main source of basal ganglia neurons (Brand and Rakic, 1979; Sidman and Rakic, 1982), as well as the subclass of interneurons for the thalamus (Rakic and Sidman, 1969; Letinic and Rakic, 2001) and in rodents, cerebral cortex interneurons (De Carlos et al., 1996; Anderson et al., 1997a, 1999, 2001; Tamamaki et al., 1997; Chapouton et al., 1999; Lavdas et al., 1999; Wichterle et al., 1999). The GE is also a major source of forebrain oligodendrocytes (He et al., 2001; Marshall and Goldman, 2002; Ulfig et al., 2002; Rakic and Zecevic, 2003). In contrast to pyramidal cells that migrate radially from the cortical ventricular zone (Sidman and Rakic, 1973; Rakic, 1974, 1988), cortical interneurons use tangential migration to reach the developing cortex. Dlx 1 and 2 (Anderson et al., 1997a,b, 2001; Eisenstat et al., 1999; Corbin et al., 2001), Nkx 2.1 (Sussel et al., 1999; Marin et al., 2000) and Lhx 6.1 (Lavdas et al., 1999) homeobox genes, which in rodents are expressed in the GE, are thought to play a role in the differentiation and tangential migration of cortical interneurons from the ventral forebrain to the neocortex. These transcription factors have the advantage of being region specific in early development, and thus can provide the information about the site of origin for different cell types later on, after the cell migrated to different, often distant, regions.

Although much has been learned in the past few years about regulation of genesis and migration of cortical interneurons in rodents, information in primates is still scarce. In the monkey, diffuse expression of Dlx1 mRNA in layer I indicated that in primates, too, some cortical neurons have a subcortical origin (Zecevic and Rakic, 2001). However, in contrast to rodents, where the majority if not all cortical interneurons originate in the GE, in primates a large number of cortical interneurons seem to originate in the cortical ventricular (VZ) and subventricular (SVZ) zones (Rakic and Zecevic, 2001; Zecevic and Rakic, 2001; Letinic et al., 2002). Therefore, in human brains, several proliferative zones, the GE, cortical VZ and SVZ, as well as the subpial granular layer (SGL), are potential sources of layer I neurons in later fetal development. The migratory routes of neurons from these different sources are more complex than has been recognized. For example, the SGL that is much more prominent and last longer in primates, has been suggested to serve as a conduit for late generated neurons coming from the olfactory region (Meyer and Goffinet, 1998; Meyer and Wahle, 1999; Zecevic and Rakic, 2001). Moreover, here we observed a dynamic trafficking in both directions, as judged by the orientations of the leading processes of migratory interneurons, at the subcortical–cortical junction, and in the subventricular and intermediate zones of human fetal brains.

Primate Cajal–Retzius cells are among the first cells to appear in the primordial plexiform layer (PPL) (Meyer and Goffinet, 1998; Zecevic et al., 1999; Meyer et al., 2000; Zecevic and Rakic, 2001) but their exact origin is still under investigation. Further-more, genesis of layer I neurons in primates last throughout the entire course of cortical neurogenesis, which complicates interpretation of their pedigree (Zecevic and Rakic, 2001). The role of Cajal–Retzius cells in secreting the glycoprotein Reelin, which is necessary for normal migration and regular layering of cortical plate neurons, is well described (D’Arcangelo et al., 1995, 1997; Ogawa et al., 1995; Del Rio et al., 1997). It appears likely that several classes of Cajal–Retzius neurons exist, and that they have different origins (Zecevic and Rakic, 2001; Meyer et al., 2002).

The normal development of layer I in humans may have considerable clinical importance, since interneurons play a substantial role in the function of the cerebral cortex, and their impairment has long been suspected in some psychiatric disorders. Cajal–Retzius cells also have an important role during normal cortical development. Their impairment is related to developmental disorders of the human cerebral cortex with disturbed neural migration, resulting in lissencephaly and mental retardation (Rakic and Caviness, 1995; Clark et al., 1997; Hong et al., 2000). Additional roles for Cajal–Retzius cells, such as in schizophrenia (Impagnatiello et al., 1998), or in repair of brain lesions (Super et al., 1997) have been suggested.

In this study, using in situ hybridization and cell specific markers, we focus on various sources and possible migration routes of preplate and layer I neurons. Preliminary results have been previously reported (Rakic and Zecevic, 2001).

Material and Methods

Tissue

Human embryos and fetuses in three age groups, 5–9 gestational weeks, (g.w.; n = 8), 11–13 g.w. (n = 4) and 17–29 g.w. (n = 5), were obtained from legal abortions/autopsies with approval of the Ethics Committee (Table 1). Brain tissue was fixed in 4% paraformaldehyde, cryoprotected by immersion in 30% sucrose, and frozen in isopentane cooled to −70°C. Frozen embryos or brain blocks of fetuses were serially sectioned in coronal, sagittal or horizontal planes at 14 μm and used for either in situ hybridization, single and double immunohistochemical experiments, or TUNEL in situ assay.

Immunohistochemistry

Antibodies (Table 2) that label interneurons (GABA, calbindin — CB, calretinin — CalR) were used in combination with antibodies specific for GE-generated cells, Dlx family and Nkx2.1. Two antibodies were used to label DLX: anti-DLX2 and pan-DLX (DLL) antibody that recognizes DLX1, 2, 5 and 6. Anti-Reelin antibodies were used to label Cajal–Retzius neurons and determine their relationship to cortical interneurons. The SMI-38 antibody reacts with the nonphosphorylated epitope of heavy neurofilaments and labels neuronal cell bodies, dendrites and some thick axons. In our human material this antibody selectively labeled Cajal– Retzius cells in preplate/layer I. Anti-Golli antibody was used to label preplate neurons (Landry et al., 1998; Hevner et al., 2001; Tosic et al., 2002). Antibodies to β-III-Tubulin and microtubule-associated protein 2 (MAP2) were used to label undifferentiated and mature neurons, respectively. Antibody to proliferating cell nuclear antigen (PCNA) was used to assess cell proliferation. The list of antibodies is presented in Table 2.

Primary antibodies were left overnight at room temperature or for 3 days at 4°C (DLX2, Dll and NKX2.1). The primary antibodies were diluted in blocking serum containing 1% bovine serum albumin, 5% normal goat serum, and 0.5% Triton X-100. Immunoreaction was revealed either by the immunoperoxidase method, employing the avidin–biotin-horseradish peroxidase (HRP) complex (ABC kit, Vector Labs, Burlingame, CA) and aminoethercarbazol (AEC, Vector Labs) or 3,3′-diaminobenzidine (DAB, Sigma) as chromogens, or by immunofluorescence method, using fluorescein or rhodamine-conjugated secondary antibodies (Jackson ImmunoResearch Labs). The sections were lightly counterstained with cresyl violet or bis-benzamide and viewed with the Leica light microscope or confocal laser-scanning fluorescence microscope (Carl Zeiss, LSM 410). Controls were processed without the primary antibodies.

In Situ Hybridization

Clones and Probes

Dlx2 and Nkx2.1 mouse clones were obtained from the laboratory of Dr John Rubenstein, University of California San Francisco. Sense and anti-sense riboprobes were generated by in vitro transcription, cloned downstream of SP6, T7 or T3 promoters, with the corresponding RNA polymerase in the presence of [35S]UTP (NEN Life Science Products Inc.).

Radioactive in situ hybridization of frozen human CNS sections was performed following the protocol previously described by Gall and Isackson (Gall and Isackson, 1989) with minor modifications. Frozen sections were first treated with 0.75% glycine in 0.1 M phosphate buffer (PB) for 5 min, rinsed with 0.1 M PB and transferred for 10 min into a solution of 0.25% acetic anhydride in 0.1 M triethanolamine, pH 8. Sections were then prehybridized for 2 h at 60°C in a 40% formamide, 10% dextran sulfate, 1 × Denhardts’s solution (0.02% Ficoll, 0.02% polyvinylpyrrolidone, 0.1% RNase-free bovine serum albumine), 1 mg/ml yeast tRNA, 10 mM DTT, 1 mg/ml denatured and sheared salmon sperm DNA and 4 × SSC. Sections were hybridized overnight at 60°C in pre-hybridization buffer containing 35S-labeled sense or antisense RNA probe (1200 Ci/mmol). After hybridization, sections were washed in 4 × SSC (30 min, 60°C twice) and digested for 30 min at 37°C in a RNase buffer (500 m NaCl, 1 mM EDTA, 10 mM Tris-HCL, pH 7.5) containing 20 μg/ml RNase A. At the end, sections were washed in 2 × SSC (30 min, four times), 0.5 × SSC (30 min, 60°C, twice), 0.1 × SSC (15 min, twice) and DEPC-treated water (15 min). Sections were exposed to β-max-hypersensitive film (Amersham Pharmacia Biotech Inc.) for 5 days and then dipped into NTB2 liquid emulsion (Eastman Kodak) for 5 weeks. Control sections hybridized with 35S-sense probe were not labeled above background level.

TUNEL In Situ Method

TUNEL in situ method, described previously (Rakic and Zecevic, 2000), was used to determine whether programmed cell death is responsible for SGL disappearance. In short, biotinylated dUTP molecules (Roche Molecular Biochemicals), catalyzed by terminal transferase (TdT) enzyme (Roche Molecular Biochemicals), were incorporated into nuclear DNA and visualized with a peroxidase standard Vectastain ABC kit (Vector Labs) and AEC (Vector Labs). In a negative control, the TdT enzyme step was excluded.

Definition of Anatomical Terms Used in This Study

Preplate or primordial plexiform layer (PPL) is a layer composed of afferent and efferent fibers and scattered neurons above the ventricular zone at the embryonic stages of development (5–8 g.w.; Carnegie stages 14–22). By 8 g.w., PPL is split by the emerging cortical plate establishing layer I above and the subplate layer below the cortical plate (Zecevic, 1993; Marin-Padilla, 1998). The subpial granular layer (SGL) is a transient, primate-specific layer of cells on top of layer I, under the pia, present in humans from 11 to 29 g.w. (Brun, 1965; Gadisseux et al., 1992; Meyer and Wahle, 1999; Zecevic and Rakic, 2001).

Results

The origin of first neurons in the wall of the cerebral vesicle

The cerebral wall at 5 g.w. embryo (Carnegie stages 14–15), the earliest stage examined here, consisted only of two developmental zones: the proliferative ventricular zone (VZ), and a cell sparse PPL layer above the VZ. Neurons labeled with antibodies to β-III-tubulin, MAP-2, Golli and calbindin were the first cells observed in the neocortical PPL (Fig. 1A — red, BE). These early, radially oriented neocortical neurons in the VZ were not co-labeled with ventral transcription factors, NKX2.1 or DlX (Fig. 1G,L), which suggested that these initial neurons have a cortical origin. In the same embryonic brain, numerous NKX2.1-positive (+) cells were present in the ventricular zone of the ventral telencephalon (preoptic area) and diencephalon (rostral hypothalamus) (Fig. 1A — green, F). NKX2.1 expression had a sharp dorsal border towards the cerebral cortex. In the region of the future hypothalamus, detached neurons contained Nkx2.1 (Fig. 1F), showing a developmental gradient in the prosencephalon even at these initial stages of development. In contrast to this, the expression of DLX family transcription factors was spread more dorsally, into the developing cerebral cortex, up to the region of the paleocortex (Fig. 1H). In the paleocortical PPL, the density of DLX labeled nuclei was several times bigger than in the VZ (Fig. 1K). This was consistent with the finding that in both ventral telencephalon and ventral diencephalon, DLX was mainly expressed in the outer, non-proliferative zone (Fig. 1HJ). DLX-expressing cells were also present outside the nervous tissue (Fig. 1H).

Dual Origin of Preplate Interneurons

In a slightly more advanced stage (6–7 g.w. or Carnegie stages 19–20), but still before the formation of the cortical plate, the PPL contains numerous GABA+ (Zecevic and Milosevic, 1997) and GAD65/67+ (Meyer et al., 2000) cells. To characterize further these cells, and to look for their possible origin and migratory routes, we performed in situ hybridization and immunohistochemistry, using probes and antibodies to ventral transcription factors DLX2/DLL and NKX2.1 along with antibodies to the neuronal cell markers.

First we tested the hypothesis that GABA+ interneurons in the embryonic PPL originate from the GE, as has been shown initially in rodents. Both NKX2.1 and DLX proteins were widely distributed through the GE and dorsal telencephalon (Fig. 2A,B). In the cortex, double labeled NKX2.1/GABA and DLX2/GABA cells were observed in the PPL (Fig. 2C,D). They were horizontally oriented and parallel to the pia. It was noted that DLX2+ cells were predominantly present in the lower part of the PPL (Fig. 2D), where the pioneer neurons were described (Meyer et al., 2000). These neurons will reside in the subplate layer once the emerging cortical plate divides the preplate.

The embryonic GE also contains Dlx2 mRNAs, as revealed by in situ hybridization, mainly in the proliferative ventricular zone (VZ) of the GE. However, spreading of the signal dorsally into the cortical VZ was noted (Fig. 3A). In a slightly later stage (9 g.w.), Dlx2 mRNA expression was confined to the proliferative zones in the GE (Fig. 3B) and mostly to the non-proliferative cortical layers, although a weaker signal was also observed in the cortical ventricular and subventricular zones (Fig. 3C). The specificity of in situ hybridization in the GE is illustrated with anti-sense and sense Dlx2 probes on adjacent sections of the embryonic brain (Fig. 3E,F).

Numerous interneurons labeled with antibodies to GABA, calbindin (CB) and calretinin (CalR) could be observed in the GE and cortical wall in sagittal sections of the 6–7 g.w. embryo. A band of immunolabeled cells connected the GE with the rostral or ventral telencephalic wall (the paleocortex) (Fig. 2EG) (Zecevic et al., 1999). Labeled cells had typical migratory morphologies. Importantly, the orientation of the leading processes suggested a two-directional tangential migration through the striato-cortical junction, with some cells directed towards the paleocortex and others towards the GE-VZ (Fig. 2GJ). The tangential stream of cells was hardly ever observed between the GE and the caudal cortex. These results are consistent with the notion that cortical interneurons originate in the GE even before cells start migrating to form the cortical plate.

In order to study whether some forebrain interneurons originate in the cortical VZ at the pre-cortical plate stage, we studied in more detail the expression of ventral transcription factors in human embryos.

In the rostral telencephalon, a strong expression of NKX2.1 was observed in the VZ lining the lateral ventricle, continuously from the GE to the cortical VZ (Figs 2A and 5A). These VZ cells were not co-labeled with either GABA or MAP2 and probably represent immature ventricular cells (Fig. 5AC). Similar densely packed, radial cells were present in the caudal GE (Fig. 5F). In the caudal cortex less densely packed NKX2.1, DLX2 and GABA immunopositive cells formed columns in the cortical and hippocampal VZ that alternated with patches without immuno-reactivity (Fig. 2A,B; Fig. 5D,E,GI). Some of these cells reached the PPL and acquire neuronal fate as shown by their double labeling with MAP2 (Fig. 5D,E).

As NKX2.1 and DLX2 have nuclear localization, it was difficult to assess the orientation of labeled cells in respect to the pia in order to estimate their migrating directions in the cerebral wall. However, double labeling with the early neuronal marker, β-III-tubulin, revealed bipolar morphology and radial orientation of these cells. Some of DLX2/β-III-tubulin and NKX2.1/β-III-tubulin cells spanned the entire thickness of the cerebral wall and appeared to be attached to both the ventricular and pial surface. This orientation suggests their radial migration between the VZ and upper regions of the telencephalic wall, and their cortical origin (Fig. 5G,H). However, in both the cerebral wall and in the GE, some cells were detached, with thick leading process directed either towards the VZ, or towards the pia. The single labeled GABAergic cells were also oriented for the most part radially within the telencephalic wall (Fig. 5I). Only an occasional horizontal GABAergic cell could be observed in the cortical VZ (Fig. 5J), indicating that GABAergic interneurons use both tangential and radial migration through the cortical wall. Double labeled NKX2.1/GABA and DLX2/GABA cells were horizontally oriented and parallel to the pia in the preplate layer (Fig. 2C,D). However, similar tangentially oriented cells were not present in the cortical VZ. This finding is in contrast to their expected tangential migration from the GE, and together with the fact that radially oriented DLX2 and NKX2.1 neurons were present in the cortical VZ, makes it very likely that some inter-neurons in humans originate in the cortical VZ at embryonic stages.

Does a Subpopulation of Cajal–Retzius Cells Originate in the GE?

In order to study whether some Cajal–Retzius cells originate in the GE, co-localization analysis was done with Reelin, a marker of Cajal–Retzius cells, and ventral transcription factors, NKX2.1 and DLX. In this study we considered Reelin+ cells situated under the pia and above the layer of interneurons to be Cajal–Retzius cells (Fig. 2E,F). During the preplate period, Reelin+ cells were numerous in the olfactory bulb and rostral cortex (Fig. 6A,B). The DLX2+ and NKX2.1+ cells also populated both regions (Fig. 6B), but these ventral telencephalic transcription factors were not expressed by Reelin+ cells. In contrast, in the caudal cortex, some of the Reelin+ cells, in the lower part of the preplate, expressed DLX2 (Fig. 6C). Additionally, the double-labeled Reelin/DLX2 cells were found in the mantle zone of the caudal GE (Fig. 6D). Reelin+ cells were also observed in the band that connects the GE to the caudal cortex (Fig. 2F). These results suggest a region-dependent origin of Reelin+ Cajal–Retzius cells — rostrally located cells presumably originated from the local cortical VZ, while a small population of caudally observed cells probably came from the GE, most likely the LGE, as we never observed Reelin/NKX2.1+ cells.

During midgestation, Reelin+ cells were also found in the globus pallidus, co-labeled with NKX2.1 (Fig. 6E), but we do not consider these to be prospective Cajal–Retzius cells. A stream of NKX2.1 cells connected the globus pallidus and ventral cortex, where it broke up into rostral and caudal branches (Fig. 6F). At the same place, Reelin was expressed in the cells under the pia, but not in the surrounding NKX2.1+ cells (Fig. 6F inset). This indicates that the stream of NKX2.1+ cells that connects the GE and the rostral telencephalon does not carry Cajal–Retzius cells. It also suggests that Reelin protein might have a role in the development of the human globus pallidus. As supporting evidence, Reelin’s role was proposed in the compartmentalization of the striatum in the postnatal rat (Nishikawa et al., 1999).

Contribution of the Subpial Granular Layer to the Cell Population of Layer I

At 11 g.w., layer I became more complex since at this time the subpial granular layer (SGL) starts emerging on the ventral brain surface (Fig. 7B,C). In sections immunolabeled with calretinin, the first cells that were forming the SGL could be observed to spread from the olfactory region to the nearby ventral cortex. After 13 g.w., the SGL covered the entire cortical surface of the forebrain. In accord with previous studies, the SGL consisted of small GABAergic cells and large Reelin+ Cajal–Retzius cells (Fig. 7D) (Meyer and Goffinet, 1998; Meyer and Wahle, 1999; Zecevic and Rakic, 2001).

At midgestation many GABAergic cells in the SGL were Dlx+ (Fig. 7E; see below). At the same time Reelin+ Cajal–Retzius cells within the SGL were labeled exceptionally well with antibody to neurofilament protein, SMI 38, but hardly ever co-expressed NKX2.1 and DLX transcription factors (Fig. 7F,G). A gradient of Reelin+ cells was present in the SGL, with numerous Reelin+ cells in the ventral cortex, close to the olfactory bulb, and their gradual tapering off towards the lateral and dorsal cortex (Fig. 7H,I). At the same time, a limited number of neurons were produced in the same ventral region of the SGL, as judged by PCNA labeling at midgestation (Fig. 7J). Other parts of the SGL did not show cell proliferation, similar to what has been reported for the Macaque monkey (Zecevic and Rakic, 2001). By the time of SGL disappearance (27–29 g.w.), only rare cells were labeled with the TUNEL method, consistent with the notion that apoptosis is not the main mechanism of SGL disappearance (Fig. 7K, Rakic and Zecevic, 2000).

Cortical Interneurons at Midgestation

At midgestation (17–22 g.w.), the in situ signal for Dlx2 and Nkx2.1 mRNAs was observed in the GE, cortical/hippocampal VZ and SVZ, and developing cortical layers (Fig. 4). A thin line of the Dlx2 and Nkx2.1 mRNA signal in the cortical VZ, with uneven border towards the surrounding tissue, is consistent with the presence of these mRNAs in the cortical VZ cells and the radial migration of these cells towards the overlying cortex (Fig. 4A,D). A strong Dlx2.1 and Nkx2.1 mRNAs signal connected the GE and the olfactory region (Fig. 4B,C).

Immunocytochemistry revealed DLX (antibody DLL) and NKX2.1+ cells in all cortical layers, including layer I and the SGL (Fig. 7EG). Almost all small GABAergic cells of the SGL can be labeled with DLX antibody (Fig. 7E), and less often with NKX2.1. In contrast, in deeper cortical layers, the NKX2.1+ cells were approximately four times more numerous than Dlx cells (not shown). This would suggest that interneurons from MGE migrate mainly through the SVZ, the intermediate zone and deep cortical layers.

A large rostral SVZ represents an additional source of cortical interneurons at midgestation (Fig. 8). Numerous vertical cell bands spread from the SVZ towards the overlying cortical plate. These cells, labeled by interneuron markers (CalR, GABA), seem to be migrating in continuous chains along cell bands. In the upper part of the SVZ, cells were aligned along fiber tracts that form a grid-like structure, crossing each other’s trajectory at right angles. The majority of these cells were also interneurons (Fig. 8E,F). CalR and GABA+ cells had the morphology of radially or tangentially migrating neurons, with larger leading and thinner trailing processes. Similar to the striato-cortical junction in embryonic stages, adjacent immunolabeled cells were often oriented in opposite directions, either towards the cortical plate or cortical VZ, if radially positioned. If tangentially positioned, the cells were oriented either medially or laterally (Fig. 8E,F).

Numerous cells in the cortical SVZ were immunoreactive to DLX2 and NKX2.1 antibodies and also seen in the cell bands that stretched from the SVZ towards the cerebral cortex (Fig. 8C,D). It was, however, not possible to show whether these cells actually reach layer I, which is several thousand microns away in midgestational fetal brains. Intensive proliferation in the SVZ was shown by PCNA labeling (Fig. 8B), but it was difficult to distinguish between the proliferation of progenitor cells that arrived from the GE and proliferation of cells that have SVZ origin.

Discussion

The present results demonstrate that both cell types in the human cortical layer I, interneurons and Cajal–Retzius cells, may have multiple origins and complex migratory routes. In contrast to rodents, neurons are continuously added to layer I during two-thirds of the gestational period in monkey, as shown by tritiated thymidine labeling studies (Zecevic and Rakic, 2001). Similar mechanisms are likely to be present in human brains, where additional proliferative centers (SVZ and SGL) overlap throughout most of the prolonged fetal development in human (Letinic et al., 2002). Indeed, total cell number (glia and neurons) increases in the fetal human brain, first rapidly from 13 to 22 g.w., and then more slowly up to term (Badsberg Samuelsen et al., 2003).

Embryonic Sources of Layer I Interneurons

A subset of layer I neurons of the human cerebral cortex express NKX and DLX families of transcription factors that are characteristic for the GE. In situ studies corroborated this finding. At the same time, NKX2.1 and DLX expressing cells could be labeled with known markers of cortical interneurons (GABA, CalR or CB). This is similar to what has been reported for rodents (De Carlos et al., 1996; Anderson et al., 1997a,b, 2001; Tamamaki et al., 1997; Pearlman et al., 1998; Chapouton et al., 1999; Lavdas et al., 1999; Marin and Rubenstein, 2001), implying that the GE origin of cortical interneurons is well conserved during evolution. However, in the early embryonic stages, before the cortical plate develops, a sub-population of interneurons that express ventral transcription factors (NKX2.1 and DLX2) was observed to be radially oriented in the cortical wall. One explanation of this result is that DLX2 and NKX2.1 cells migrate first tangentially from the GE to the cortical VZ, and then radially through the cortical wall. This has been described in rodents (De Carlos et al., 1996; Anderson et al., 2001). However, retroviral labeling of VZ/SVZ cells in slice preparation of the fetal human forebrain show that these cells divide several times before starting radial migration to their final cortical destination, arguing for their cortical origin in humans (Letinic et al., 2002). In support of this finding, we observed that mRNAs signal and immunolabeling of ventral transcription factors, spread dorsally, to embryonic cortical areas. At the same time, tangentially oriented NKX2.1 and DLX2+ cells were lacking in the embryonic proliferative zone of the forebrain. Taken together, these results suggest that a subpopulation of NKX2.1 and DLX2+ cells present in the cortex, were not coming from the GE. Rather, a more likely possibility is that ventral transcription factors expand dorsally in human brains, labeling also regions of the cortical VZ. Thus, in human brain, ventral transcription factors, DLX and NKX2.1, may not be specific only for GE derived cells. This conclusion is supported both by our in situ results, and by the recent study of human fetal organotypic slice cultures (Letinic et al., 2002). These authors observed that in human brain only one-third of Dlx1,2-expressing cortical inter-neurons have GE origin, whereas two-thirds have neocortical origin and migrate radially to the overlying cortical plate. This predominance of radial migration over tangential is reflected in the prominent radial organization of the human cerebral cortex during mid-gestation before it become obscured by formation of layers and elaboration of connectivity (McKinstry et al., 2002). The constant migration of later born neurons from the VZ/SVZ towards the cerebral cortex is reflected in the size of the subplate layer, which reaches its peak in the second half of gestation in human brain (Kostovic et al., 2002).

Thus, in contrast to rodents, the much larger human brain has additional sources of cortical interneurons that include cortical VZ/SVZ and later on, the SGL.

Bi-directional Neuronal Migration in the Telencephalon

In addition to radial migration towards the pia, as typically observed in the cortical wall, and transverse migration from GE to the neocortex, many cells were heading in the ‘opposite’ direction. Thus, a possibility that some cortical interneurons, originating from GE, reach preplate layer by tangential migration, and subsequently descend into the cortical wall migrating radially towards the VZ, should also be considered. This type of migration has been observed in live preparation of mouse forebrain by multiphoton microscopy (Ang et al., 2003). This direction of migration would be an additional explanation for the radially oriented DLX2, NKX2.1 or GABA+ cells observed in the embryonic cerebral wall. Thus, both interneurons coming from the cortical VZ, and a subpopulation of cells descending from the layer I, might be radially oriented in the cortical wall. Furthermore, at midgestation, in the subventricular and the intermediate zone, numerous cortical interneurons had their leading processes directed towards the cortical VZ. Similar to this, in the GE, some cells were directed towards the GE-VZ (Fig. 2J for example). This observation, based only on morphology of cells in human fetal brain, recently was well documented in rodent brain slices and named ‘ventricle-directed migration’ (Nadarajah et al., 2002). The suggestion that the VZ is supplying specific layer information to interneurons, similar to information that pyramidal neurons obtain from the cortical VZ, is worth further investigation (Nadarajah et al., 2002).

A similar phenomenon was present in the striato-cortical junction, where some GABA and CalR interneurons had their leading processes directed towards the GE, as if migrating from the cortex to the GE. Numerous neuronal fibers (calretinin, calbindin, neurofilament proteins-antibodies SMI31 and SMI32) or non-neuronal processes (vimentin, GFAP) were also crossing through the same striato-cortical junction from the early embryonic stages, possibly participating in migration through the junction (Zecevic et al., 1999). Our observation is consistent with the report on embryonic rat slice culture that showed the ‘inward migration’ into the developing striatum (Hamasaki et al., 2001).

However, the possibility that cells go forward and then retract for some steps before going forward again, observed by time lapse microscopy for glial cells in organotypic slice cultures (Kakita and Goldman, 1999), cannot be eliminated in our study on frozen sections of human fetal brain.

The Subpial Granular Layer and the Subpopulation of Cajal–Retzius Cells Contain Ventral Transcription Factors

A transient subpial granular layer (SGL), described first in primates (Brun, 1965; Gadisseux et al., 1992; Meyer and Goffinet, 1998; Meyer and Wahle, 1999; Zecevic and Rakic, 2001) contains granular GABAergic interneurons and Cajal–Retzius cells. SGL cells originate from the SVZ close to the olfactory bulb, migrate to the brain surface and give rise to subpial granular cells (Gadissaux et al., 1992; Meyer et al., 1998; Zecevic and Rakic, 2001). We have now established that SGL cells were coming from the olfactory region, and that the majority of them express DLX. Either the Dlx cells first migrated from the GE to the olfactory region, or this transcription factor was expressed there from the beginning, since both DLX and NKX2.1+ cells were present in the olfactory region from embryonic stages. In addition, at midgestation, the mRNA signals for the two transcription factors were expressed as a continuous band between the GE and the olfactory region, stressing their connection. At the time when the subpial granular layer starts forming, these cells could spread over the forebrain surface as part of this layer.

The SGL disappeared by 27–29 g.w., probably by inward migration of its cells, as was described in the monkey (Zecevic and Rakic, 2001). Rare TUNEL+ cells in this layer show that apoptosis cannot be sufficient for the observed quick removal of this layer (Spreafico et al., 1999; Rakic and Zecevic, 2000).

The existence of several subpopulations of Cajal–Retzius cells indicated on the basis of different antigen expression (Lavdas et al., 1999; Meyer et al., 2000, 2002; Zecevic et al., 1999; Zecevic and Rakic, 2001), was confirmed in this study by their expression of different transcription factors. Region specific transcription factors can provide the information about the site of origin for different cell types later on, after the cell migrated to distant regions. In the case of Reelin+ Cajal–Retzius cells, the presence of DLX in subpopulation of these cells speaks in favor that this subpopulation has an LGE origin. Furthermore, bipolar Reelin+ neurons with migratory morphologies were present in the caudal GE of embryonic brains, whereas a stream of Reelin+ cells was observed in continuum from the GE to the neocortex and ventral telencephalon. At the same time the observed lack of radially migrating Reelin+ cells in the cortical wall has been suggested to be due to the fact that Cajal–Retzius cells express Reelin only after migrating to their position under the pia (Meyer and Wahle, 1999). However, present results show that migrating cells in the GE can express Reelin. This is consistent with the view that a subpopulation of Cajal–Retzius cells, traditionally believed to be of cortical origin, could have a subcortical origin from the GE. Recently, similar conclusions have been reached in monkey (Zecevic and Rakic, 2001) and human (Meyer et al., 2002) developing forebrain.

The lack of NKX2.1 expression in Cajal–Retzius cells observed in this study is consistent with the results in rat, where genetically labeled cells transplanted in the MGE migrated to the MZ, but did not express Reelin (Wichterle et al., 2001). Another transcription factor characteristic for MGE, Lhx6, is expressed in some Cajal–Retzius cells (Lavdas et al., 1999). At the same time, the expression of Tbr1 in many Cajal–Retzius cells argues that a larger subpopulation of these cells comes from the cortical ventricular zone (Hevner et al., 2001). As was discussed above, differences between human and other mammals are likely to exist. In primates the GE might have a larger repertoire of different cell types compared with rodents. Supporting evidence is that, in humans, the GE also contributes neurons to the thalamic nuclei (Sidman and Rakic, 1973; Letinic and Kostovic, 1997; Letinic and Rakic, 2001), and oligodendrocyte progenitors to the forebrain (Ulfig et al., 2002; Rakic and Zecevic, 2003).

In summary, the prolonged genesis and multiple sites of origin of cortical interneurons and Cajal–Retzius cells in the human brain may have important clinical implications. In preterm infants intracerebral hemorrhage in the GE or the periventricular leukomalacia in the near-by subventricular zone (Volpe, 2001; Back et al., 2001), can destroy a large number of interneurons and oligodendrocytes destined for the cerebral cortex. Depletion of cells in this region during development might have serious consequences, as both cortical interneurons and Cajal–Retzius cells are implicated in schizophrenia (Benes et al., 1991; Akbarian et al., 1995; Impagnatiello et al., 1998; Lewis, 2000) and bipolar disorder (Knable, 1999). The GE is also the site where oligoprogenitors originate (Ulfig et al., 2002; Rakic and Zecevic, 2003), and where thalamo-cortical and corticofugal fibers meet and influence each other’s target finding (Molnar and Blackemore, 1995; Metin and Godement, 1996; Ulfig et al., 2000). Thus, it is not surprising that in preterm infants lesions in this region result in some common birth defects (Larroche, 1964; Gadisseux et al., 1992; Volpe, 2001; Back et al., 2001; Ulfig, 2001).

Notes

The authors are very thankful to Dr Kent Morest and Dr Pasko Rakic for their valuable comments on this article, to Dr Daniel Geschwind for his help with in situ experiments, and to Drs M. Ogawa, A. Goffinet, J. Rubenstein, S. Anderson, Y. Kohtz and A. Campagnoni for their generous gifts of antibodies and probes. This study was supported by grants from NIH-NS 41 489-02 and the National Multiple Sclerosis Society, RG-3083-A-1.

Corresponding author: Dr Nada Zecevic, Department of Neuroscience, University of Connecticut School of Medicine, Farmington, CT 06030, USA. Email: nzecevic@neuron.uchc.edu.

Table 1

Human embryos and fetuses used in this study

Period Case Age (g.w.) Carnegie stage Embedding Plane 
Embryonic E1 14–15 frozen sagittal 
 E2 16 frozen frontal 
 E3 17 frozen sagittal 
 E4 18–19 frozen sagittal 
 E5 18–19 frozen sagittal 
 E6 18–19 frozen sagittal 
 E7 18–19 frozen horizontal 
 E8 18–19 paraffin sagittal 
Early fetal F1 – frozen sagittal 
 F2 11 – frozen sagittal 
 F3 13 – frozen sagittal 
 F4 13 – frozen frontal 
Midgestation F5 17 – frozen frontal 
 F6 17 – frozen frontal 
 F7 22 – frozen sagittal 
 F8 25 – frozen sagittal 
 F9 27–28 – frozen sagittal 
Period Case Age (g.w.) Carnegie stage Embedding Plane 
Embryonic E1 14–15 frozen sagittal 
 E2 16 frozen frontal 
 E3 17 frozen sagittal 
 E4 18–19 frozen sagittal 
 E5 18–19 frozen sagittal 
 E6 18–19 frozen sagittal 
 E7 18–19 frozen horizontal 
 E8 18–19 paraffin sagittal 
Early fetal F1 – frozen sagittal 
 F2 11 – frozen sagittal 
 F3 13 – frozen sagittal 
 F4 13 – frozen frontal 
Midgestation F5 17 – frozen frontal 
 F6 17 – frozen frontal 
 F7 22 – frozen sagittal 
 F8 25 – frozen sagittal 
 F9 27–28 – frozen sagittal 
Table 2

Antibodies used in this study

Cell type Name Host Dilution Manufacture 
Interneurons anti-calbindin mouse IgG 1:1000 Sigma, St Louis, MO, USA 
 anti-calretinin rabbit IgG 1:5000 SWant, Switzerland 
 anti-GABA mouse IgG 1:100 Sigma, St Louis, MO, USA 
 anti-GABA rabbit IgG 1:1000 IncStar, Stillwater, M, USA 
Cajal–Retzius cells anti-Reelin mouse IgG, clones 142 and CR50 1:500 Gift Drs Goffinet & Ogawa 
 SMI38 mouse IgG 1:1000 Sternberger 
Preplate cells anti-Golli rabbit IgG 1:100 Gift Dr A. Campagnoni 
GE-originating cells anti-Dlx2 rabbit IgG 1:50 CeMines, Evergreen, CO, USA 
 anti-DLL rabbit IgG 1:40 Gift Dr Y. Kohtz 
MGE-originating cells anti-Nkx2.1 rabbit IgG 1:500 Biopat Immunotechnologies, Italy 
Neurons anti-MAP2 mouse IgG 1:200 Sigma, St Louis, MO, USA 
 anti-β-III-Tubulin mouse IgG 1:400 Sigma, St Louis, MO, USA 
Proliferating cells anti-PCNA mouse IgG 1:50 DAKO, Carpinteria, CA, USA 
Cell type Name Host Dilution Manufacture 
Interneurons anti-calbindin mouse IgG 1:1000 Sigma, St Louis, MO, USA 
 anti-calretinin rabbit IgG 1:5000 SWant, Switzerland 
 anti-GABA mouse IgG 1:100 Sigma, St Louis, MO, USA 
 anti-GABA rabbit IgG 1:1000 IncStar, Stillwater, M, USA 
Cajal–Retzius cells anti-Reelin mouse IgG, clones 142 and CR50 1:500 Gift Drs Goffinet & Ogawa 
 SMI38 mouse IgG 1:1000 Sternberger 
Preplate cells anti-Golli rabbit IgG 1:100 Gift Dr A. Campagnoni 
GE-originating cells anti-Dlx2 rabbit IgG 1:50 CeMines, Evergreen, CO, USA 
 anti-DLL rabbit IgG 1:40 Gift Dr Y. Kohtz 
MGE-originating cells anti-Nkx2.1 rabbit IgG 1:500 Biopat Immunotechnologies, Italy 
Neurons anti-MAP2 mouse IgG 1:200 Sigma, St Louis, MO, USA 
 anti-β-III-Tubulin mouse IgG 1:400 Sigma, St Louis, MO, USA 
Proliferating cells anti-PCNA mouse IgG 1:50 DAKO, Carpinteria, CA, USA 
Figure 1 (left).

In human embryos NKX2.1 and DLX transcription factors are expressed in a complementary way and confined to the ventral prosencephalon. (A) The drawing of the sagittally cut 5-week-old embryo (Carnegie stages 14–15) shows the expression of neuronal markers (red) and NKX2.1 transcription factor (green). The neurons in the marginal zone of the prospective neocortex express (B) β-III-Tubulin, (C) MAP2, (D) Golli and (E) calbindin (CB). Note several radially oriented cells in the VZ (arrows) labeled with neuronal markers. (F) NKX2.1 is expressed in the VZ of the preoptic area and rostral part of the hypothalamus, whereas (G) the future dorsal cortex is NKX2.1-negative. At the same time, DLX family of transcription factors is mostly expressed in the PPL of the ventral prosencephalon (H), in preoptic area (I), basal ganglia (J), very sparsely in the paleocortex (K) and not yet in the neocortex (L). Neurons in the PPL, on I, J, K, L, are labeled with the antibody to β-III-tubulin, The broken line on I, J, K, L represents the border between nervous and non-nervous tissue. Note the DLX expression in leptomeninges (L) and outside the CNS (arrowheads, H). BG, basal ganglia; Cx, cortex; nCx, neocortex; pCx, paleocortex; H, hypothalamus; M, mesencephalon; P, prosencephalon; POA, preoptic area; PPL, preplate layer; R, rhombencephalon; S, septum; Th, thalamus; VZ, ventricular zone; V, ventricle; β-III-T — beta-III-tubulin. Scale bars = 100 μm (BG, IJ); and 300 μm (H).

Figure 1 (left).

In human embryos NKX2.1 and DLX transcription factors are expressed in a complementary way and confined to the ventral prosencephalon. (A) The drawing of the sagittally cut 5-week-old embryo (Carnegie stages 14–15) shows the expression of neuronal markers (red) and NKX2.1 transcription factor (green). The neurons in the marginal zone of the prospective neocortex express (B) β-III-Tubulin, (C) MAP2, (D) Golli and (E) calbindin (CB). Note several radially oriented cells in the VZ (arrows) labeled with neuronal markers. (F) NKX2.1 is expressed in the VZ of the preoptic area and rostral part of the hypothalamus, whereas (G) the future dorsal cortex is NKX2.1-negative. At the same time, DLX family of transcription factors is mostly expressed in the PPL of the ventral prosencephalon (H), in preoptic area (I), basal ganglia (J), very sparsely in the paleocortex (K) and not yet in the neocortex (L). Neurons in the PPL, on I, J, K, L, are labeled with the antibody to β-III-tubulin, The broken line on I, J, K, L represents the border between nervous and non-nervous tissue. Note the DLX expression in leptomeninges (L) and outside the CNS (arrowheads, H). BG, basal ganglia; Cx, cortex; nCx, neocortex; pCx, paleocortex; H, hypothalamus; M, mesencephalon; P, prosencephalon; POA, preoptic area; PPL, preplate layer; R, rhombencephalon; S, septum; Th, thalamus; VZ, ventricular zone; V, ventricle; β-III-T — beta-III-tubulin. Scale bars = 100 μm (BG, IJ); and 300 μm (H).

Figure 2 (right).

NKX 2.1 and DLX 2 transcription factors in the interneurons of the cortical PPL and ganglionic eminence (GE) in the late human embryonic period (Carnegie stages 19–20). (A, B) Drawings of the coronal/horizontal sections of the human embryonic brain at the level of the GE, with the distribution of NKX2.1 and DLX2 transcription factors in the telencephalon (green). Note that NKX2.1+ and DLX2+ nuclei are arranged in patches through the width of the caudal neocortical ventricular zone. (C) In an NKX2.1/GABA double-labeled section, horizontally oriented double-labeled cells are present in the cortical PPL (arrow). Higher magnification of one of these cells is presented in the inset. (D) Double-labeled DLX2/GABA cells are also observed in the lower portion of the PPL (arrow). (E) Reelin and GABA label different cell populations in both neocortex and the GE. Arrow shows the orientation of majority of GABA+ leading processes. (F) Higher magnification of immunolabeled cells between the GE and neocortex. Arrows indicate possible routes of their migration based on orientation of leading processes. Double white arrow shows predominant orientation of GABA+ cells (rostrally and ventrally), while red arrow indicates possible migration of Reelin+ cells towards the caudal cortex. (G–H) Different tangential migration patterns in embryonic forebrain. (G) Bi-directional tangential migration of interneurons labeled with calretinin (CalR) through the striato-cortical junction. (H, I, J) Higher magnification of migrating cells. Leading processes indicated by the arrows, are directed either towards the GE or towards the paleocortex. Open arrow on H points in direction of the neocortex. OB, olfactory bulb; GE, ganglionic eminence; LV, lateral ventricle; hipp, hippocampus; Th, thalamus; H, hypothalamus; III, third ventricle; PPL, preplate; VZ, ventricular zone; r, rostral; l, lateral; d, dorsal; pCx, paleocortex; nCx, neocortex. Scale bars = 100 μm (C); 50 μm (C inset, D, HJ); 300 μm (E); and 200 μm (F, G).

Figure 2 (right).

NKX 2.1 and DLX 2 transcription factors in the interneurons of the cortical PPL and ganglionic eminence (GE) in the late human embryonic period (Carnegie stages 19–20). (A, B) Drawings of the coronal/horizontal sections of the human embryonic brain at the level of the GE, with the distribution of NKX2.1 and DLX2 transcription factors in the telencephalon (green). Note that NKX2.1+ and DLX2+ nuclei are arranged in patches through the width of the caudal neocortical ventricular zone. (C) In an NKX2.1/GABA double-labeled section, horizontally oriented double-labeled cells are present in the cortical PPL (arrow). Higher magnification of one of these cells is presented in the inset. (D) Double-labeled DLX2/GABA cells are also observed in the lower portion of the PPL (arrow). (E) Reelin and GABA label different cell populations in both neocortex and the GE. Arrow shows the orientation of majority of GABA+ leading processes. (F) Higher magnification of immunolabeled cells between the GE and neocortex. Arrows indicate possible routes of their migration based on orientation of leading processes. Double white arrow shows predominant orientation of GABA+ cells (rostrally and ventrally), while red arrow indicates possible migration of Reelin+ cells towards the caudal cortex. (G–H) Different tangential migration patterns in embryonic forebrain. (G) Bi-directional tangential migration of interneurons labeled with calretinin (CalR) through the striato-cortical junction. (H, I, J) Higher magnification of migrating cells. Leading processes indicated by the arrows, are directed either towards the GE or towards the paleocortex. Open arrow on H points in direction of the neocortex. OB, olfactory bulb; GE, ganglionic eminence; LV, lateral ventricle; hipp, hippocampus; Th, thalamus; H, hypothalamus; III, third ventricle; PPL, preplate; VZ, ventricular zone; r, rostral; l, lateral; d, dorsal; pCx, paleocortex; nCx, neocortex. Scale bars = 100 μm (C); 50 μm (C inset, D, HJ); 300 μm (E); and 200 μm (F, G).

Figure 3 (left).

Expression of Dlx2 mRNA in human developing telencephalon during embryonic and early fetal period using radioactive in situ hybridization. (A, B) The signal for Dlx2 mRNA, visualized with 35S-labeled antisense probe, in the ganglionic eminence (GE) of 6-week-old embryos. Note: cortex in (B) is wrinkled by artifact. (C) Higher magnification of the 9-week-old embryonic cortex: Dlx2 signal is predominantly observed in the cortical layers with differentiated cells (SP, CP, LI). (D) Adjacent section is stained with cresyl violet to show the cytoarchitecture with developing cortical plate (CP — between lines). (E) Intracellular localization of the Dlx2 mRNA in the GE was detected using anti-sense probe. (F) No signal in the same region was observed with the sense probe. Cx, cerebral cortex; VZ, ventricular zone; SVZ, subventricular zone; IZ, intermediate zone; SP, subplate layer; CP, cortical plate; LI, layer I; r, rostral; l, lateral; d, dorsal. Scale bar = 5 mm (A, B), 50 μm (C, D), 10 μm (E, F).

Figure 3 (left).

Expression of Dlx2 mRNA in human developing telencephalon during embryonic and early fetal period using radioactive in situ hybridization. (A, B) The signal for Dlx2 mRNA, visualized with 35S-labeled antisense probe, in the ganglionic eminence (GE) of 6-week-old embryos. Note: cortex in (B) is wrinkled by artifact. (C) Higher magnification of the 9-week-old embryonic cortex: Dlx2 signal is predominantly observed in the cortical layers with differentiated cells (SP, CP, LI). (D) Adjacent section is stained with cresyl violet to show the cytoarchitecture with developing cortical plate (CP — between lines). (E) Intracellular localization of the Dlx2 mRNA in the GE was detected using anti-sense probe. (F) No signal in the same region was observed with the sense probe. Cx, cerebral cortex; VZ, ventricular zone; SVZ, subventricular zone; IZ, intermediate zone; SP, subplate layer; CP, cortical plate; LI, layer I; r, rostral; l, lateral; d, dorsal. Scale bar = 5 mm (A, B), 50 μm (C, D), 10 μm (E, F).

Figure 4 (right).

Expression of Dlx2 and Nkx2.1 mRNAs in human developing telencephalon during midgestation as seen by radioactive in situ hybridization. (A) In the frontal sections through both hemispheres of 17 g.w. fetus, Dlx2 mRNA is expressed mainly in the GE-SVZ and, as a weaker signal, in the cortical VZ with radially directed bands (arrow). In situ signal is also present in the cerebral cortex. (B) On sagittally cut section at 25 g.w. a strong Dlx2 mRNA signal is found from GE to the olfactory region and in the hippocampus. The frontal pole of the brain is to the left of the field; occipital pole to the right. (C, D) The expression of Nkx2.1 mRNA is revealed in two sagittal sections through the rostral part of the telencephalon at 22 g.w. (C) in the more medial section, Nkx2.1 mRNA is express in the neocortical SVZ, the cerebral cortex, and in the GE, spreading towards the olfactory region, whereas (D) in the more lateral section, signal is observed also in the proliferative zones of GE and in the basal ganglia. Arrows point to the bands of the neocortical subventricular zone. Frontal pole of the brain is to the left. VZ, ventricular zone; SVZ, subventricular zone; Cx, cortex; GE, ganglionic eminence; nc, nucleus caudatus; p, putamen; ic, internal capsule; LV, lateral ventricle; Th, thalamus; hipp, hippocampus; OR, olfactory region; gp, globus pallidus. Scale bars = 5 mm (AD).

Figure 4 (right).

Expression of Dlx2 and Nkx2.1 mRNAs in human developing telencephalon during midgestation as seen by radioactive in situ hybridization. (A) In the frontal sections through both hemispheres of 17 g.w. fetus, Dlx2 mRNA is expressed mainly in the GE-SVZ and, as a weaker signal, in the cortical VZ with radially directed bands (arrow). In situ signal is also present in the cerebral cortex. (B) On sagittally cut section at 25 g.w. a strong Dlx2 mRNA signal is found from GE to the olfactory region and in the hippocampus. The frontal pole of the brain is to the left of the field; occipital pole to the right. (C, D) The expression of Nkx2.1 mRNA is revealed in two sagittal sections through the rostral part of the telencephalon at 22 g.w. (C) in the more medial section, Nkx2.1 mRNA is express in the neocortical SVZ, the cerebral cortex, and in the GE, spreading towards the olfactory region, whereas (D) in the more lateral section, signal is observed also in the proliferative zones of GE and in the basal ganglia. Arrows point to the bands of the neocortical subventricular zone. Frontal pole of the brain is to the left. VZ, ventricular zone; SVZ, subventricular zone; Cx, cortex; GE, ganglionic eminence; nc, nucleus caudatus; p, putamen; ic, internal capsule; LV, lateral ventricle; Th, thalamus; hipp, hippocampus; OR, olfactory region; gp, globus pallidus. Scale bars = 5 mm (AD).

Figure 5.

NKX2.1 and DLX2 transcription factors and GABA in the ventricular zone of the telencephalon in the 6–7-week-old human embryo (Carnegie stages 19–20). On the drawing, four boxed areas show where subsequent fields are. (A) Boxed area 1: NKX2.1 immunoreactivity is restricted to the VZ and it is seen continuously from GE to the anterior cerebral cortex. The midline is labeled with an asterisk. (B) MAP2 immunoreactivity is confined to the non-proliferative zones, PPL and mantle layer. (C) Merged image shows complementary expression of these two markers. (D) Boxed area 2: in the caudal cortex NKX2.1/MAP2 double-labeled cells are present in the PPL (arrows), whereas cells with nuclear expression of the NKX2.1 are dispersed through the VZ in patches. (E) Boxed area 3: in the hippocampus more cells show nuclear expression of NKX2.1 in the VZ, whereas few cells in the PPL are neurons (MAP2+) with NKX2.1 expression (arrows). (F) Boxed area 4: increased density of NKX2.1+ cells in the ventricular zone of the caudal GE as compared with the neocortex. (G) In the caudal cortex cells labeled with neuronal marker β-III-tubulin span the whole thickness of the VZ and express DLX2 in the perikaryon (arrow). (H) Rare double-labeled NKX2.1/β-III-tubulin cell oriented radially in the VZ of the rostral cortex (arrow). (I, J) GABAergic cells in the VZ and the PPL of the caudal cortex. Both radial and horizontal (arrow) orientations of these cells can be seen. GE, ganglionic eminence; LV, lateral ventricle; III, hird ventricle; Cx, cortex; cCx, caudal cortex; cGE, caudal ganglionic eminence; rCx, rostral cortex; PPL, preplate; VZ, ventricular zone. Scale bars = 300 μm (AC) and 50 μm (DJ).

Figure 5.

NKX2.1 and DLX2 transcription factors and GABA in the ventricular zone of the telencephalon in the 6–7-week-old human embryo (Carnegie stages 19–20). On the drawing, four boxed areas show where subsequent fields are. (A) Boxed area 1: NKX2.1 immunoreactivity is restricted to the VZ and it is seen continuously from GE to the anterior cerebral cortex. The midline is labeled with an asterisk. (B) MAP2 immunoreactivity is confined to the non-proliferative zones, PPL and mantle layer. (C) Merged image shows complementary expression of these two markers. (D) Boxed area 2: in the caudal cortex NKX2.1/MAP2 double-labeled cells are present in the PPL (arrows), whereas cells with nuclear expression of the NKX2.1 are dispersed through the VZ in patches. (E) Boxed area 3: in the hippocampus more cells show nuclear expression of NKX2.1 in the VZ, whereas few cells in the PPL are neurons (MAP2+) with NKX2.1 expression (arrows). (F) Boxed area 4: increased density of NKX2.1+ cells in the ventricular zone of the caudal GE as compared with the neocortex. (G) In the caudal cortex cells labeled with neuronal marker β-III-tubulin span the whole thickness of the VZ and express DLX2 in the perikaryon (arrow). (H) Rare double-labeled NKX2.1/β-III-tubulin cell oriented radially in the VZ of the rostral cortex (arrow). (I, J) GABAergic cells in the VZ and the PPL of the caudal cortex. Both radial and horizontal (arrow) orientations of these cells can be seen. GE, ganglionic eminence; LV, lateral ventricle; III, hird ventricle; Cx, cortex; cCx, caudal cortex; cGE, caudal ganglionic eminence; rCx, rostral cortex; PPL, preplate; VZ, ventricular zone. Scale bars = 300 μm (AC) and 50 μm (DJ).

Figure 6 (left).

Reelin expression and origin of Reelin+ Cajal–Retzius cells. (A, B) Embryonic Reelin- and DLX2-expressing cells are present in the olfactory bulb and rostral cortex where they represent two different, non-overlapping cell populations. However, few Reelin/DLX2+ horizontally oriented cells are observed in the PPL of the caudal cortex (arrow, C) and LGE (arrow, D). (E) At midgestation, Reelin+ cells in the globus pallidus express NKX2.1. (F) A stream of NKX 2.1+ cells is observed between the GP and the subpial zone of both, the olfactory region and the hypothalamus. Although Reelin is expressed in the olfactory region (arrow), it is not present in the NKX2.1+ cells coming from the GP, as shown in inset. OB, olfactory bulb; rCx, rostral cortex; cCX, caudal cortex; LGE, lateral ganglionic eminence; GP, globus pallidus; OR, olfactory region; H, hypothalamus; PPL, primordial plexiform layer. Scale bars = 300 μm (A); 100 μm (B, D, F); and 50 μm (C, E, F inset).

Figure 6 (left).

Reelin expression and origin of Reelin+ Cajal–Retzius cells. (A, B) Embryonic Reelin- and DLX2-expressing cells are present in the olfactory bulb and rostral cortex where they represent two different, non-overlapping cell populations. However, few Reelin/DLX2+ horizontally oriented cells are observed in the PPL of the caudal cortex (arrow, C) and LGE (arrow, D). (E) At midgestation, Reelin+ cells in the globus pallidus express NKX2.1. (F) A stream of NKX 2.1+ cells is observed between the GP and the subpial zone of both, the olfactory region and the hypothalamus. Although Reelin is expressed in the olfactory region (arrow), it is not present in the NKX2.1+ cells coming from the GP, as shown in inset. OB, olfactory bulb; rCx, rostral cortex; cCX, caudal cortex; LGE, lateral ganglionic eminence; GP, globus pallidus; OR, olfactory region; H, hypothalamus; PPL, primordial plexiform layer. Scale bars = 300 μm (A); 100 μm (B, D, F); and 50 μm (C, E, F inset).

Figure 7 (top right).

Human subpial granular layer (SGL). (A) Drawing of a sagittal section of the 11 g.w. brain. (B) Olfactory bulb and adjacent ventral cortex labeled with anti-calretinin antibody. (C) Higher magnification of the boxed area in B depicts the beginning of SGL formation in the vicinity of the olfactory bulb. (D) Double labeling in the SGL at 13 g.w. reveals two types of cells: small GABAergic granular cells (green) and large Reelin+ Cajal–Retzius cells (red). (E) At 22 g.w., double labeling with GABA and DLL (pan-DLX) antibodies shows that all small SGL granular cells are GABAergic and many of them are also DLL+ (arrow). (F) Cajal–Retzius cells in the SGL are labeled with SMI38 antibody, and small granular cells with DLL. (G) Reelin (red) and NKX2.1 (green, arrow) are not co-localized in the same cells in the SGL. NKX2.1 is expressed by few small cells. (H) Drawing of a frontal section through the hemisphere of 17 g.w. fetus. Arrowheads point to the ventral cortex where pictures presented in I, J and K are taken. The sizes of the arrowheads represent the decreasing gradient of Reelin+ cells from olfactory bulb and ventral cortex towards the latero-dorsal cortex. (I) Reelin+ Cajal–Retzius cells in the SGL. (J) Cell proliferation assessed by the antibody to PCNA is observed only in the SGL of the ventral cortex. (K) Rare apoptotic cell (arrow) in the SGL detected by TUNEL in situ method; Cx, cortex; LV, lateral ventricle; OB, olfactory bulb; Th, thalamus; mc, mesencephalon; rc, rombencephalon; VZ, ventricular zone; SP, subplate zone; CP, cortical plate; SGL, subpial granular layer; BG, basal ganglia; NCx, neocortex; PCx, paleocortex; CalR, calretinin; PCNA, proliferating cell nuclear antigen. Scale bars = 200 μm (B, C); 100 μm (D, IK); and 50 μm (EG).

Figure 7 (top right).

Human subpial granular layer (SGL). (A) Drawing of a sagittal section of the 11 g.w. brain. (B) Olfactory bulb and adjacent ventral cortex labeled with anti-calretinin antibody. (C) Higher magnification of the boxed area in B depicts the beginning of SGL formation in the vicinity of the olfactory bulb. (D) Double labeling in the SGL at 13 g.w. reveals two types of cells: small GABAergic granular cells (green) and large Reelin+ Cajal–Retzius cells (red). (E) At 22 g.w., double labeling with GABA and DLL (pan-DLX) antibodies shows that all small SGL granular cells are GABAergic and many of them are also DLL+ (arrow). (F) Cajal–Retzius cells in the SGL are labeled with SMI38 antibody, and small granular cells with DLL. (G) Reelin (red) and NKX2.1 (green, arrow) are not co-localized in the same cells in the SGL. NKX2.1 is expressed by few small cells. (H) Drawing of a frontal section through the hemisphere of 17 g.w. fetus. Arrowheads point to the ventral cortex where pictures presented in I, J and K are taken. The sizes of the arrowheads represent the decreasing gradient of Reelin+ cells from olfactory bulb and ventral cortex towards the latero-dorsal cortex. (I) Reelin+ Cajal–Retzius cells in the SGL. (J) Cell proliferation assessed by the antibody to PCNA is observed only in the SGL of the ventral cortex. (K) Rare apoptotic cell (arrow) in the SGL detected by TUNEL in situ method; Cx, cortex; LV, lateral ventricle; OB, olfactory bulb; Th, thalamus; mc, mesencephalon; rc, rombencephalon; VZ, ventricular zone; SP, subplate zone; CP, cortical plate; SGL, subpial granular layer; BG, basal ganglia; NCx, neocortex; PCx, paleocortex; CalR, calretinin; PCNA, proliferating cell nuclear antigen. Scale bars = 200 μm (B, C); 100 μm (D, IK); and 50 μm (EG).

Figure 8 (bottom right).

Cortical subventricular zone (SVZ) at midgestation (17–22 g.w.). (A) Radial cell bands from the VZ and SVZ are directed towards the cerebral cortex (arrows). (B) Numerous PCNA+ nuclei in the SVZ bands suggest that cells are proliferating while on their migrating path. (C) Numerous DLL labeled cells in the SVZ. (D) NKX2.1+ nuclei (green) in the cortical SVZ cells co-labeled with neural marker MAP2 (red) in cell bands (arrows) spanning from the SVZ to the cortex. Inset: higher magnification of NKX2.1/MAP2+ neuron in the cell band. (E) All calretinin (CalR) labeled interneurons are co-labeled with GABA (F). Arrows point to the direction of a leading process (yellow — radial, blue — horizontal migration); pia is towards the upper right corner of the photograph; Scale bars = 500 μm (A); and 50 μm (BG).

Figure 8 (bottom right).

Cortical subventricular zone (SVZ) at midgestation (17–22 g.w.). (A) Radial cell bands from the VZ and SVZ are directed towards the cerebral cortex (arrows). (B) Numerous PCNA+ nuclei in the SVZ bands suggest that cells are proliferating while on their migrating path. (C) Numerous DLL labeled cells in the SVZ. (D) NKX2.1+ nuclei (green) in the cortical SVZ cells co-labeled with neural marker MAP2 (red) in cell bands (arrows) spanning from the SVZ to the cortex. Inset: higher magnification of NKX2.1/MAP2+ neuron in the cell band. (E) All calretinin (CalR) labeled interneurons are co-labeled with GABA (F). Arrows point to the direction of a leading process (yellow — radial, blue — horizontal migration); pia is towards the upper right corner of the photograph; Scale bars = 500 μm (A); and 50 μm (BG).

References

Akbarian S, Kim JJ, Potkin SG, Hagman JO, Tafazzoli A, Bunney WE Jr, Jones EG (
1995
) Gene expression for glutamic acid decarboxylase is reduced without loss of neurons in prefrontal cortex of schizophrenics.
Arch Gen Psychiatry
 
52
:
258
–266.
Anderson SA, Eisenstat DD, Shi L, Rubenstein JL (
1997
) Interneuron migration from basal forebrain to neocortex: dependence on Dlx genes.
Science
 
278
:
474
–476.
Anderson SA, Qiu M, Bulfone A, Eisenstat DD, Meneses J, Pedersen R, Rubenstein JL (
1997
) Mutations of the homeobox genes Dlx-1 and Dlx-2 disrupt the striatal subventricular zone and differentiation of late born striatal neurons.
Neuron
 
19
:
27
–37.
Anderson SA, Mione M, Yun K, Rubenstein JL (
1999
) Differential origins of neocortical projection and local circuit neurons: role of Dlx genes in neocortical interneuronogenesis.
Cereb Cortex
 
9
:
646
–654.
Anderson SA, Marin O, Horn C, Jennings K, Rubenstein JL (
2001
) Distinct cortical migrations from the medial and lateral ganglionic eminences.
Development
 
128
:
353
–363.
Ang ESBC, Haydar TF, Gluncic V, Rakic P (
2003
) Four-dimensional migratory coordinates of GABAergic interneurons in the developing mouse cortex.
J Neurosci
 
23
:
5805
–5815.
Back SA, Luo NL, Borenstein NS, Levine JM, Volpe JJ, Kinney HC (
2001
) Late oligodendrocyte progenitors coincide with the developmental window of vulnerability for human perinatal white matter injury.
J Neurosci
 
21
:
1302
–1312.
Badsberg Samuelsen G, Bonde Larsen K, Bogdanovic N, Laursen H, Gream N, Larsen JF, Pakkenberg B (
2003
) The changing number of cells in the human fetal forebrain and its subdivisions: a stereological analysis.
Cereb Cortex
 
13
:
115
–122.
Benes FM, McSparren J, Bird ED, SanGiovanni JP, Vincent SL (
1991
) Deficits in small interneurons in prefrontal and cingulate cortices of schizophrenic and schizoaffective patients.
Arch Gen Psychiatry
 
48
:
996
–1001.
Brand S, Rakic P (
1979
) Genesis of the primate neostriatum: [3H]thymidine autoradiographic analysis of the time of neuron origin in the rhesus monkey.
Neuroscience
 
4
:
767
–778.
Brun, A (
1965
) The subpial granular layer of the foetal cerebral cortex in man. Its ontogeny and significance in congenital cortical malformations.
Acta Pathol Microbiol Scand
  (Suppl. 179)
13
:
1
–98.
Chapouton P, Gartner A, Gotz M (
1999
) The role of Pax6 in restricting cell migration between developing cortex and basal ganglia.
Development
 
126
:
5569
–5579.
Corbin JG, Nery S, Fishell G (
2001
) Telencephalic cells take a tangent: non-radial migration in the mammalian forebrain.
Nat Neurosci
 
4
(Suppl.):
1177
–1182.
Clark GD, Mizuguchi M, Antalffy B, Barnes J, Armstrong D (
1997
) Predominant localization of the LIS family of gene products to Cajal– Retzius cells and ventricular neuroepithelium in the developing human cortex.
J Neuropathol Exp Neurol
 
56
:
1044
–1052.
De Carlos JA, Lopez-Mascaraque L, Valverde F (
1996
) Dynamics of cell migration from the lateral ganglionic eminence in the rat.
J Neurosci
 
16
:
6146
–6156.
D’Arcangelo G, Miao GG, Chen SC, Soares HD, Morgan JI, Curran T (
1995
) A protein related to extracellular matrix proteins deleted in the mouse 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 monoclonal antibody.
J Neurosci
 
17
:
23
–31.
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 development of hippocampal connections.
Nature
 
385
:
70
–74.
Eisenstat DD, Liu JK, Mione M, Zhong W, Yu G, Anderson SA, Ghattas I, Puelles L, Rubenstein JL (
1999
) DLX-1, DLX-2, and DLX-5 expression define distinct stages of basal forebrain differentiation.
J Comp Neurol
 
414
:
217
–237.
Gadisseux JF, Goffinet AM, Lyon G, Evrard P (
1992
): The human transient subpial granular layer: an optical, immunohistochemical, and ultrastructural analysis.
J Comp Neurol
 
324
:
94
–114.
Gall CM and Isackson PJ (
1989
) Limbic seizures increase neuronal production of messenger RNA for nerve growth factor.
Science
 
245
:
758
–761.
Grasbon-Frodl EM, Nakao N, Lindvall O, Brundin P (
1996
) Phenotypic development of the human embryonic striatal primordium: a study of cultured and grafted neurons from the lateral and medial ganglionic eminences.
Neuroscience
 
73
:
171
–183.
Hamasaki T, Goto S, Nishikawa S, Ushio Y (
2001
) Early-generated preplate neurons in the developing telencephalon: inward migration into the developing striatum.
Cereb Cortex
 
11
:
474
–484.
He W, Ingraham C, Rising L, Goderie S, Temple S (
2001
) Multipotent stem cells from the mouse basal forebrain contribute GABAergic neurons and oligodendrocytes to the cerebral cortex during embryogenesis.
J Neurosci
 
21
:
8854
–8862.
Hevner RF, Shi L, Justice N, Hsueh Y-P, Sheng M, Smiga S, Bulfone A, Goffinet AM, Rubenstein, JLR (
2001
) Tbr1 regulates differentiation of the preplate and layer 6.
Neuron
 
29
:
353
–366.
Hong SE, Haung DT, Shugart YY, Al Shahwan S, Grant PE, Hourihane JOB, Martin NDT, Walsh CA (
2000
) Autosomal recessive lissencephaly with cerebellar hypoplasia (LCH) is associated with human reelin gene mutations.
Nature Genetics
 
26
:
93
–96.
Impagnatiello F, Guidotti AR, Pesold C, Dwivedi Y, Caruncho H, Pisu MG, Uzunov DP, Smalheiser NR, Davis JM, Pandey GN, Pappas GD, Tueting P, Sharma RP, Costa E (
1998
) A decrease of reelin expression as a putative vulnerability factor in schizophrenia.
Proc Natl Acad Sci USA
 
95
:
15718
–15723.
Kakita A, Goldman JE (
1999
) Patterns and dynamics of SVZ cell migration in the postnatal forebrain: monitoring living progenitors in slice preparations.
Neuron
 
23
:
461
–472.
Knable MB (
1999
) Schizophrenia and bipolar disorder: findings from studies of the Stanley Foundation Brain Collection.
Schizophr Res
 
39
:
149
–152.
Kostovic I, Judas M, Rados M, Hrabac P (
2002
) Laminar organization of the human fetal cerebrum revealed by histochemical markers and magnetic resonance imaging.
Cereb Cortex
 
12
:
563
–544.
Landry CF, Pribyl TM, Ellison JA, Givogri MI, Kampf K, Campagnoni C and Campagnoni AT (
1998
). Embryonic expression of the myelin basic protein gene: identification of a promoter region that targets transgene expression to pioneer neurons.
J Neurosci
 
18
:
7315
–7327.
Larroche JC (
1964
) Hemorragies cerebrales intraventriculaires chez le premature. I partie: Anatomie et Physiopathologie.
Biol Neonate
 
7
:
26
–56.
Lavdas AA, Grigoriou M, Pachnis V, Parnavelas JG (
1999
) The medial ganglionic eminence gives rise to a population of early neurons in the developing cerebral cortex.
J Neurosci
 
19
:
7881
–7888.
Letinic K, Kostovic I (
1997
) Transient fetal structure, the gangliothalamic body, connects telencephalic germinal zone with all thalamic regions in the developing human brain.
J Comp Neurol
 
384
:
373
–395.
Letinic K, Rakic P (
2001
) Telencephalic origin of human thalamic GABAergic neurons.
Nat Neurosci
 
4
:
931
–936.
Letinic K, Zoncu R, Rakic P (
2002
) Origin of GABAergic neurons in the human neocortex.
Nature
 
417
:
645
–649.
Lewis DA (
2000
) GABAergic local circuit neurons and prefrontal cortical dysfunction in schizophrenia.
Brain Res Rev
 
31
:
270
–276.
Marin O, Rubenstein JL (
2001
) A long, remarkable journey: tangential migration in the telencephalon.
Nat Rev Neurosci
 
2
:
780
–790.
Marin O, Anderson SA, Rubenstein JL (
2000
) Origin and molecular specification of striatal interneurons.
J Neurosci
 
20
:
6063
–6076.
Marin-Padilla M (
1998
) The Cajal–Retzius cell and the development of the neocortex.
Trends Neurosci
 
21
:
64
–71.
Marshall CA, Goldman JE (
2002
) Subpallial Dlx2-expressing cells give rise to astrocytes and oligodendrocytes in the cerebral cortex and white matter.
J Neurosci
 
22
:
9821
–9830.
Metin C, Godement P (
1996
) The ganglionic eminence may be an intermediate target for corticofugal and thalamocortical axons.
J Neurosci
 
16
:
3219
–3235.
McKinstry RC, Mathur A, Miller JH, Ozean A, Snyder AZ, Schefft GL, Almli CR, Shiran SI, Conturo TE, Neil JJ (
2002
) Radial organization of developing preterm human cerebral cortex revealed by non-invasive water diffusion anisotropy MRI.
Cereb Cortex
 
12
:
1237
–1243.
Meyer G, Goffinet AM (
1998
) Prenatal development of reelin-immunoreactive neurons in the human neocortex.
J Comp Neurol
 
397
:
29
–40.
Meyer G, Wahle P (
1999
) The paleocortical ventricle is the origin of reelin-expressing neurons in the marginal zone of the fetal human neocortex.
Eur J Neurosci
 
11
:
3937
–3944.
Meyer G, Soria JM, Martinez-Galan JR, Martin-Clemente B, Fairen A (
1998
) Different origins and developmental histories of transient neurons in the marginal zone of the fetal and neonatal rat cortex.
J Comp Neurol
 
397
:
493
–518.
Meyer G, Schaaps JP, Moreau L, Goffinet AM (
2000
) Embryonic and early fetal development of the human neocortex.
J Neurosci
 
20
:
1858
–1868.
Meyer G, Perez-Garcia CG, Abraham H, Caput D (
2002
) Expression of p73 and reelin in the developing human cortex.
J Neurosci
 
22
:
4973
–4986.
Molnar Z, Blakemore C (
1995
) How do thalamic axons find their way to the cortex?
Trends Neurosci
 
18
:
389
–397.
Nadarajah B, Alifragis P, Wong RO, Parnavelas JG (
2002
) Ventricle-directed migration in the developing cerebral cortex.
Nat Neurosci
 
5
:
218
–224.
Nishikawa S, Goto S, Hamasaki T, Ogawa M, Ushio Y (
1999
) Transient and compartmental expression of the reeler gene product reelin in the developing rat striatum.
Brain Res
 
850
:
244
–248.
Ogawa M, Miyata T, Nakajima K, Yagyu K, Seike M, Ikenaka K, Yamamoto H, Mikoshiba K (
1995
) The reeler gene-associated antigen on Cajal–Retzius neurons is a crucial molecule for laminar organization of cortical neurons.
Neuron
 
14
:
899
–912.
O’Rahilly R, Muller F (
1994
) The embryonic human brain. An atlas of developmental stages. New York: Wiley Liss.
Parnavelas JG (
2000
) The origin and migration of cortical neurons: new vistas.
Trends Neurosci
 
23
:
126
–131.
Pearlman AL, Faust PL, Hatten ME, Brunstrom JE (
1998
) New directions for neuronal migration.
Curr Opin Neurobiol
 
8
:
45
–54.
Rakic P (
1974
) Neurons in rhesus monkey visual cortex: systematic relation between time of origin and eventual disposition.
Science
 
183
:
425
–427.
Rakic P (
1988
) Specification of cerebral cortical areas.
Science
 
241
:
170
–176.
Rakic P, Caviness VS Jr (
1995
) Cortical development: view from neurological mutants two decades later.
Neuron
 
14
:
1101
–1104.
Rakic P, Sidman RL (
1969
) Telencephalic origin of pulvinar neurons in the fetal human brain.
Z Anat Entwicklungsgesch
 
129
:
53
–82.
Rakic S, Zecevic N (
2000
) Programmed cell death in the developing human telencephalon.
Eur J Neurosci
 
12
:
2721
–2734.
Rakic S, Zecevic N (
2001
) NKX2.1 and DLX2 are expressed in human developing telencephalon. Soc Neurosci Abstr 468.3.
Rakic S, Zecevic N (
2003
) Early oligodendrocyte progenitor cells in the human fetal telencephalon.
Glia
 
41
:
117
–127.
Sidman RL, Rakic P (
1973
) Neuronal migration, with special reference to developing human brain: a review.
Brain Res
 
62
:
1
–35.
Sidman RL, Rakic P (
1982
) Development of the human central nervous system. Springfield, IL: CC Thomas.
Spreafico R, Arcelli P, Frassoni C, Canetti P, Giaccone G, Rizzuti T, Mastrangelo M, Bentivoglio M (
1999
) Development of layer I of the human cerebral cortex after midgestation: architectonic findings, immunocytochemical identification of neurons and glia, and in situ labeling of apoptotic cells.
J Comp Neurol
 
410
:
126
–142.
Super H, Sust PP, Soriano E (
1997
) Survival of Cajal–Retzius cells after cortical lesions in newborn mice: a possible role for Cajal–Retzius cells in brain repair.
Dev Brain Res
 
98
:
9
–14.
Sussel L, Marin O, Kimura S, Rubenstein JL (
1999
) Loss of Nkx2.1 homeobox gene function results in a ventral to dorsal molecular respecification within the basal telencephalon: evidence for a transformation of the pallidum into the striatum.
Development
 
126
:
3359
–3370.
Tamamaki N, Fujimori KE, Takauji R (
1997
) Origin and route of tangentially migrating neurons in the developing neocortical intermediate zone.
J Neurosci
 
17
:
8313
–8323.
Tosic M, Rakic S, Matthieu, JM, Zecevic N (
2002
) Identification of Golli and Myelin basic proteins in human brain during early development.
Glia
 
37
:
219
–228.
Ulfig N (
2001
) Expression of calbindin and calretinin in the human ganglionic eminence.
Pediatr Neurol
 
24
:
357
–360.
Ulfig N, Setzer M, Neudorfer F, Bohl J (
2000
) Distribution of SNAP-25 in transient neuronal circuitries of the developing human forebrain.
Neuroreport
 
6
:
1259
–1263.
Ulfig N, Briese M, Bohl J (
2002
) Expression of oligo dendrocyte-specific protein (OSP)/claudin-11 in the human fetal forebrain.
Neuro-embryology
 
1
:
48
–53.
Volpe JJ (
2001
) Neurobiology of periventricular leukomalacia in the premature infant.
Pediatr Res
 
50
:
553
–562.
Wichterle H, Garcia-Verdugo JM, Herrera DG, Alvarez-Buylla A (
1999
) Young neurons from medial ganglionic eminence disperse in adult and embryonic brain.
Nat Neurosci
 
2
:
461
–466.
Wichterle H, Turnbull DH, Nery S, Fishell G, Alvarez-Buylla A (
2001
) In utero fate mapping reveals distinct migratory pathways and fates of neurons born in the mammalian basal forebrain.
Development
 
128
:
3759
–3771.
Zecevic N (
1993
) Cellular composition of the telencephalic wall in human embryos.
Early Hum Dev
 
32
:
131
–149.
Zecevic N, Milosevic A (
1997
) Initial development of gamma-amino-butyric acid immunoreactivity in the human cerebral cortex.
J Comp Neurol
 
380
:
495
–506.
Zecevic N, Rakic P (
2001
) Development of layer I neurons in the primate cerebral cortex.
J Neurosci
 
21
:
5607
–5619.
Zecevic N, Milosevic A, Rakic S, Marin-Padilla M (
1999
) Early development and composition of the human primordial plexiform layer: an immunohistochemical study.
J Comp Neurol
 
412
:
241
–254.