The cerebral cortex originates in the thickness of thetelencephalic vesicle, in the anterior segment of the mam-malian embryo. The ventricular zone (VZ) of the dorsal partof this structure is the only germinative component wherecortical cells proliferate (Marın-Padilla, 1971, 1972, 1978;Fishell et al., 1993). Following their generation in the VZ,the cells undergo a radial gliophilic migration following aninside-out gradient to occupy their definitive locations(Rakic, 1971, 1972). The ventral part of the telencephalicvesicles is divided into the medial (MGE) and lateral (LGE)ganglionic eminences, from which the corpus striatum andpallidum of the basal telencephalon (BT) are generated(Smart and Sturrock, 1979). These germinative territoriesare separated by a prominent sulcus inside the ventricularsurface, the corticostriatal limit, which was thought torestrict cell movement (Fishell, 1995). However, recentstudies have shown that, early in development, many cells
1 To whom correspondence should be addressed. Fax: �34 (91)
generated in the ganglionic eminences migrate tangentiallyacross the corticostriatal sulcus to the neocortex (De Carloset al., 1996; Anderson et al., 1997a; Tamamaki et al., 1997).These cells express GABA and Calbindin (Anderson et al.,1997a; Tamamaki et al., 1997) and are substantially reducedin the neocortex of Dlx-1 and Dlx-2 mutant mice (Andersonet al., 1997b, 1999). Nevertheless, it still remains unclearwhether these migrating cells originate in the MGE, LGE,or both. It has been shown that some Cajal-Retzius, sub-plate (SP), and lower intermediate zone (IZ) cells originatein the MGE (Lavdas et al., 1999). Similarly, Wichterle et al.(1999) have presented evidence for a contribution of cellsfrom the MGE in cortical brain structures after graftingneuronal precursors from the MGE into the adult brain.However, Anderson et al. (2001) reported that GABA-expressing cells migrate from the MGE into the cerebralcortex through the IZ, while at later stages, they found thatLGE-derived cells migrated into the neocortex via thesubventricular zone (SVZ). Thus, it appears that both theMGE and LGE could be sources of different cortical inter-neurons.
The aim of the present study was to investigate the
cortical tangential migration pathways that originate in the585 4754. E-mail: [email protected].
ganglionic eminences, along the telencephalic rostrocaudalaxis and at different embryonic ages. As a result, we canconfirm the role of the MGE in the development of thecerebral cortex and we demonstrate that cortical cells arealso generated in the LGE at earlier stages, suggesting astrictly controlled spatiotemporal pattern of migration. Wealso partially characterized these cell populations and theirmigratory routes by using different markers.
Some preliminary aspects of this work have been pre-sented previously (Jimenez et al., 1999a,b, 2000).
MATERIALS AND METHODS
Animals. Wistar rats raised in the Cajal Institute were usedthroughout the study. Prenatal animals were obtained by cesarean
FIG. 1. The formation of the ganglionic eminences during the development of the basal telencephalon. Rat embryo brains (E12–E17) havebeen stained with acid thionin. Each panel represents one embryonic age in the horizontal plane (left) and in the coronal plane (right). Theright hemisphere of the transversally sectioned brains has been shadowed to clarify the structures. BT, basal telencephalon; Ctx, cortex;LGE, lateral ganglionic eminence; MGE, medial ganglionic eminence; ne, neuroepithelium.
section from timed pregnant dams, anesthetized with Equithesin (3ml/kg body weight). The day of insemination was defined asembryonic (E) day 0, and the first 24 h after birth as postnatal (P)day 0; pups were born on E22 (P0). All animals were handled in ahumane manner to avoid major distress.
Nissl staining. Brains obtained from E12–E17 rat embryos werethick sectioned in the horizontal and transversal planes and stainedwith acid thionin (0.05%, pH 4.5).
Computer reconstruction. In order to study the disposition ofthe ganglionic eminences and the progression and depth of theinterganglionic sulcus, three-dimensional reconstructions wereobtained from serial sections by using the computer program“Design CAD Pro 2000” (ViaGrafix, Pryor, OK).
Birthdating experiments. Pregnant rats at different gestationaldays, from E12 to E17 postcoitum, were injected intraperitoneallywith a solution containing 50 mg/kg of 5-Bromo-2�-deoxyuridine-5�-monophosphate (BrdU; Boehringer Mannheim, Indianapolis,IN). Embryos were removed by cesarean section at different timesfollowing injection (Table 1) and perfused with a saline phosphatebuffer (PBS). The brains were dissected out and immersed inCarnoy fixative for 24–48 h, dehydrated with graded ethanol, andembedded in paraffin at 60°C for 3–4 h. The brains were sectionedat 10 �m, the paraffin removed, and the sections were rehydratedand incubated overnight with mouse anti-BrdU (0.25 �g/ml; Boeh-
FIG. 2. Computer reconstruction of a coronal-sectioned E14 rat brain. The right cerebral hemisphere is represented from the midline tothe lateral surface in six different perspectives by rotating the structure every 30° on the vertical axis. (A) A view from the anterior pole;(F) From the posterior side. The gray sheet represents the external surface of the brain and the red sheet the internal (ventricular) surface.White arrowheads point to the interganglionic sulcus.
TABLE 1Timing of BrdU Injections and Death of Rats
ringer Mannheim) in PBS containing 0.1% bovine serum albumin(BSA) (Schutte et al., 1987).
Slice cultures. Embryos were removed by cesarean section, andtheir brains were dissected out in ice-cold Leivobitz L15 (Gibco/BRL) containing 10% horse serum. Tissue Chopper (Mc Ilwain)coronal sections (300 �m) were collected and transferred to thesame sterile medium. The slices were cultured at 37°C with a gasmixture of 95% O2, 5% CO2 in a Type I Rat Tail Collagen matrix(Collaborative Biomedical Products) prepared with MEM (10�) andSodium Bicarbonate in 400 ml of DMEN (Gibco/BRL), withL-glutamine, penicillin/streptomycin, and horse serum added. Inother cases, hemibrains and thick pieces of tissue were dissectedout by hand with the aid of microscissors and a pair of watchmakerforceps, and cultured in collagen matrices.
Whole embryo cultures. The technique for dissecting, han-dling, and culturing embryos was largely based on the methods andprotocols described by Cockroft (1990), having been used for a longtime in our laboratory (De Carlos et al., 1996). Briefly, pregnant ratsof known gestational ages were deeply anesthetized and theiruterine horns exposed by a longitudinal abdominal incision. Em-bryos were dissected out in a petri dish containing Hank’s balancedsolution at 37°C, under sterile conditions. The muscular uterinewall and the decidua were removed, and Reichert’s membrane wasopened and dissected apart to reveal the vascularized visceral yolksac containing the embryo. Maintaining the integrity of thevitelline arteries and veins, the yolk sac was partially broken at itsavascular side to expose the embryo attached by the umbilicalvessels. The amnion was removed and the vessels of the vitellinestalk were tucked under the tail of the embryo. Tracers wereinjected under a dissecting microscope, introducing the micropi-pette through the telencephalic vesicle in a caudorostral orienta-tion in order to approach the target structures from inside theventricle. The injected embryo was then transferred to a glassbottle containing the culture medium (heat-inactivated rat serum)and placed in an incubator device, at 34°C with continuous gassing(95% O2, 5% CO2), for 1 or 2 days.
Tracers and injection procedures. The tracers used were thefluorescent 1,1�-dioctadecyl-3,3,3�,3�,-tetramethylindocarbocyanineperchlorate (DiI; Molecular Probes) (Honig and Hume, 1986, 1989)and a biotinylated dextran amine (BDA-3000; Molecular Probes).These tracers were injected at different embryonic stages, usingthree different approaches: injections in utero; injections ex utero,and posterior culture of the whole embryo in roller bottles. Alter-natively, small DiI crystals were applied to brain slices that werethen cultured in collagen matrices as previously described. DiIdissolved in dimethylformamide (0.5%) was injected in utero, insubcortical structures of whole embryos, through a fine-tippedcrystal micropipette using a pressure system (Picospritzer; GeneralValve).
Immunocytochemistry. The following antibodies were used inthis study: anti-Calbindin (CB-D28K, Swant, 1:2500); anti-Calretinin (CR, Swant, 1:1000); anti-GABA (Sigma, 1:1000); andanti-PSA-NCAM (Mouse, IgM, DSHB, 1:5000). The secondaryantibody used for the detection of all of the primary antibodies wasImmunoPure Goat anti-Rabbit IgG Biotin conjugate (Pierce, 1:100).Localization of the antibodies was performed with the ABC method(Immuno Pure Ultra-Sensitive ABC Peroxidase Staining Kit;Pierce).
Photomicroscopy and data acquisition. Fluorescent imageswere viewed and photographed by using standard photomicros-copy. Negatives and prints were digitized by using a Kodak RFS
2035 plus film scanner (Eastman-Kodak, Rochester, NY) or imageswere recorded directly at a resolution of 1,600 dpi by using a digitalcamera (DP10; Olympus, Tokyo, Japan) attached to the micro-scope. Digitized images were assembled and color and/or contrastbalanced by using Micrografx Picture Publisher (version 8) andMicrosoft Power Point (2000) software.
Some nomenclature and technical considerations. In neocor-tical development, the first postmitotic neurons accumulate super-ficially in the neuroepithelium, immediately beneath the pialsurface, forming the preplate (De Carlos and O’Leary, 1992). Todesignate this stratum, we always use the term “preplate” (coinedby Stewart and Pearlman, 1987), although other authors refer to itas “the primordial plexiform layer” (Marın-Padilla, 1971) or the“pallial anlage” (Rickmann et al., 1977). The preplate persists untilE15 in the rat, the age at which the cortical plate (CP) cellsgenerated become postmitotic. These cells reach the preplate andcause it to split, giving rise to a superficial marginal zone (futurelayer I) and a deep subplate (Marın-Padilla, 1971). Consequently, inthe present work, we use the term preplate (PP) to refer to the firststratum that appears in the neuroepithelium and, from E15 on-ward, the term marginal zone (MZ) to designate the more superfi-cial layer.
We would also like to clarify the rationale behind the differentlabeling approaches used in this work. In our initial experiments,embryos were removed from the uterus and exposed while stillattached by their umbilical vessels, and the label was injected intothe desired structure. Subsequently, the whole embryos werecultured in roller bottles for 1 or 2 days (De Carlos et al., 1996).However, most laboratories interested in tangential migration fromthe ganglionic eminences use cultured tissue slices. We believethat our approach prevents the disruption of some putative rostro-caudal migratory routes, that would otherwise be lost in slicecultures. However, brain slices are perfectly suitable for the studyof the number of cells generated in each eminence and to analyzethe contribution of the eminence to the cerebral cortex. To performa more complete analysis and to be able to compare our resultswith those that have been reported by other authors, we carried outexperiments in tissue slices, as well as labeling in utero (in vivo)and using whole embryo cultures (in vitro).
RESULTS
Macroscopic Development of the GanglionicEminences
At early developmental stages, the wall of the telence-phalic vesicles is a continuous semicircular sheet of tissuethat shows no evidence of regional specialization. To deter-mine how the two ganglionic eminences form, we haveperformed a morphological analysis in whole brains. Thefirst sign of differentiation is the appearance of a ventraldome-shaped elevated protrusion into the ventricular cav-ity (Fig. 1A). At E12, the MGE begins to emerge, althoughthere is no trace of the LGE until E12.5, when it starts tooccupy the dorsolateral part of the BT (Fig. 1B). At E13, thewell-developed MGE occupies a more caudoventral loca-tion than the LGE, which forms in the lateral part of thebasal telencephalon (Fig. 1C). At E14, both eminences haveattained roughly the same volume as can be seen in selected
coronal sections (Fig. 1D). Indeed, it is at this stage that animportant number of ganglionic eminence-generated cellsare incorporated into the nascent cerebral cortex. From E15onward, the MGE begins to regress at the same time thatthe LGE increases in size, and by E16, the LGE occupiesalmost the complete extension of the BT, although a traceof the MGE can still be detected (Fig. 1F). Finally, at E17,there is no sign of the MGE, and both lateral and medialganglionic eminences are restructured as a consistent eleva-tion of tissue that, together with other structures, forms thebasal telencephalon (Fig. 1G).
To facilitate our anatomical study of the interganglionicsulcus, we have performed a computer reconstruction fromcoronal sections of an E14 rat embryo (Fig. 2). One cerebralhemisphere is represented from the midline to the lateralsurface from six different perspectives. At midlevels of therostrocaudal axis, both eminences appear to be similar insize, although the LGE is larger than the MGE, since theformer extends frontally to the olfactory bulb, while cau-dally, it contributes to the structure known as the caudalganglionic eminence. In fact, the first one-third of the basaltelencephalon is occupied solely by the LGE at this age(Figs. 2C and 2D). It is also interesting to note that the depthof the sulcus divides the two eminences rostrally (Figs.2A–2D, arrowheads) and could have impeded MGE-generated cells from entering into the neocortical mantlethat passes through the LGE. Indeed, the sulcus betweenthe two eminences is very deep at the rostral level, but itbecomes smoother as it progresses caudally, and disappearswhere both eminences merge to form the caudal eminence(Figs. 2E and 2F). Thus, the cortical cells that originate inthe most rostral portion of the MGE might be forced tomigrate toward the caudal axis of the brain to seek an easiermigration pathway.
Birthdating Experiments
Cell generation was studied by analyzing the incorpora-tion of BrdU into S-phase mitotic cells at different embry-onic developmental stages (see Table 1). Cells generated atE12 were located at E13 in both the VZ of the corticalneuroepithelium and the VZ of the LGE (Figs. 3A and 3B).None of these cells were found in the VZ of the MGE. In thefollowing days, the cells generated at E12 were found in thePP of the cortical neuroepithelium, the BT, and the septalarea (Figs. 3C and 3D). Two hours after the BrdU injectionat E13, labeled neuroblasts were located throughout the VZ(Fig. 3E), while at E14, these cells had migrated toward thepial surface to occupy the encephalic thickness. Further-more, the VZ and SVZ of the BT also contained labeled cells(Figs. 3F and 3G). At E15, the cells generated at E13 weremainly localized in the MZ (Fig. 3H), while at E16, theywere largely located in the subplate (SP), although some ofthem still remain in the MZ. Both cell populations delimitanother group of nonlabeled cells from the CP. In addition,some cells that heavily incorporated BrdU appeared dis-
persed along the IZ, suggesting that the first tangential cellsmigrating through the IZ might be generated at E13 (Fig. 3I).Cells generated at E14 are located along the whole ence-phalic SVZ 2 h later. These cells migrate superficially and,at E15, are located in three different areas: the SP, IZ, andVZ/SVZ (Fig. 3J). Some other cells were also seen migratingradially through the thickness of the CP (Fig. 3K). At E16,cells that labeled heavily at E14 are located in the CP, whileat E17, they occupy the lower part of the CP and SP (Fig. 3L).Two hours after the BrdU injection at E15, all the SVZ cellsare labeled and, at E16, labeled cells can be seen in threedifferent strata: the VZ/SVZ, IZ, and SP (Fig. 3M). Thenuclei of the labeled migrating cells are vertically orien-tated in relation to the pial surface in the VZ/SVZ andhorizontally in the IZ and SP (Fig. 3N). We also observedsome ascending cells (with a vertical nuclear orientation)through the CP. At E17, labeled cells from E15 occupied theentire CP, while horizontally oriented cells remained in thelower part of the IZ (Fig. 3O). Cells that incorporated BrdUat E16 and were analyzed 2 h later were located in the SVZ,whereas at E17, these cells were distributed in two differentlayers: the SVZ with their nuclei oriented vertically, andthe lower IZ with a tangential nuclear orientation (Figs. 3Pand 3Q). At E18, the labeled vertical cells in the SVZ hadmigrated to the upper part of the CP, while a horizontalmigration was maintained through the lower part of the IZ(Figs. 3R and 3S).
Although the direction of migration cannot be deter-mined by BrdU labeling, the preferred orientation of thelabeled nuclei has been shown to be a good tool to infer thedirection of migration. As such, the majority of the nucleiin the lower IZ appeared to show a horizontal disposition(Figs. 3N and 3Q), while others were rounded or evenvertically orientated. It is known that this stratum iscrossed by radially ascending cells.
In summary, BrdU cell-labeling of the rat demonstratesthe well-reported radial migration of cortical cells, suggest-ing that the tangential migrations occur between E13 andE14 all along the neocortical PP. However, cell migration inthe lower part of the neocortical IZ occurs during a broaderperiod of time, beginning at E13 and with a peak ofgeneration occurring between E15 and E16.
Different Tangential Migratory Cell Populations
A detailed study of the tangential migratory cell popula-tions from the two eminences was undertaken at differentembryonic ages by using three different approaches: (i) DiIinjections into target tissue explants and subsequent cul-ture; (ii) single DiI or BDA injections into the BT of ex uteroembryos (in toto) that were then cultured in roller bottlesfor 1 or 2 days; and (iii) in utero BDA injections into the BTof living embryos.
As a result of these studies, two populations of migratingcells were identified in the BT. The first population origi-nated in the VZ of the LGE early in development (E12.5–
159Tangential Migration in Neocortical Development
FIG. 3. Neurogenesis: BrdU injections in rats. Labeled cells at E13 after injection at E12 (A, B). (B) Higher magnification of boxed area in(A). Most cells generated at E12 are located at E13 in the ventricular zone of the cortical neuroepithelium and in the ventricular zone ofthe LGE. (C, D) At E14, after injection at E12, labeled cells are mainly in the PP. Box in (C) is enlarged in (D). Above the labeled cells inthe PP, some erythrocytes stained with thionin appear in the arachnoid space. Two hours after BrdU injection at E13, labeled cells arelocated throughout the VZ and SVZ (E), while at E14, these cells migrate toward the pial surface occupying the whole encephalic thickness,including the VZ and SVZ of the basal telencephalon (F, G; box in F is enlarged in G). Cells generated at E13 are mainly localized in theMZ at E15 (H), while at E16, they are mostly located in the SP (I). At E15, cells generated at E14 are situated in the SP, IZ, and VZ/SVZ (J),although some cells migrate radially through the thickness of the CP (K, enlarged from box in J). At E17, cells labeled at E14 mainly occupythe lower part of the CP and the SP (L). After BrdU injection at E15, cells are localized in three different strata at E16: VZ/SVZ, IZ, and SP(M), the nuclari orientation of the migrating cells is vertical in relation to the pial surface in the VZ/SVZ and horizontal in the IZ and SP(N). At E17, the cells labeled at E15 occupy the entire CP, while horizontally oriented cells remain in the lower part of the IZ (O). Cellsvisualized at E17 after injection at E16 are mainly distributed in two different layers: in a vertical disposition in the SVZ and, with atangential orientation, in the lower IZ (P, Q). White arrows in (N) and (Q) point to the preferred orientation of the cells. Insets in (Q) showan enlarged view of the nuclear orientation. At E18, the vertical labeled cells in the SVZ have migrated to reach the upper part of the CP,while the horizontal migration through the lower part of the IZ is maintained (R, S). Images are coronal sections; dorsal is up and lateralis left in (A–C, E-M, O, P, R, S); dorsal is right in (D, N, Q). (A–D) and (L–S) are counterstained with thionin. az, arachnoid zone; BT, basaltelencephalon; CP, cortical plate; LGE, lateral ganglionic eminence; LIZ, lower intermediate zone; MZ, marginal zone; MGE, medialganglionic eminence; ne, neuroepithelium; PP, preplate; SVZ, subventricular zone; v, ventricle; VZ, ventricular zone. Scale bars are: 250 �min (M); 200 �m in (A, C, F); 100 �m in (B, I, J, L, O, P); 50 �m in (E, H); 25 �m in (D, K, N, S); and 15 �m in (G, Q).
E13), from where they migrated radially toward the pialsurface and later tangentially through the cortical PP (Fig.4A). One day later, cells generated in the SVZ of the LGEmigrated tangentially toward both the PP and the IZ,although at this early developmental stage some cellsoccupied the VZ and SVZ (Fig. 4B). The second populationwas generated in the MGE slightly later (E14) and migratedtangentially through the PP and the lower part of thecortical IZ (Figs. 4C and 4D). Each population of cells thatemerged from the two eminences split into two separateclusters of neurons as they entered into the cortical neuro-epithelium, although sometimes they seemed to enterrandomly. Later on, all tangentially migrating cells occu-pied the PP or MZ, and the lower IZ. Initially, each cellpopulation originated in the VZ and SVZ of both emi-nences. However, at E16, the MGE begins to disappear, andfrom E17 onward, the tangential migratory cells that passed
through the cortical mantle came from only one structure,the SVZ of the BT.
In our previous work on tangential migrations (De Carloset al., 1996), we described the migration of cells from theLGE, but we did not address the possibility that cells mightmigrate from the MGE. More recently, several groups haveobserved a migratory population of cells that emerge fromthis structure (Anderson et al., 1997; Tamamaki et al.,1997; Lavdas et al., 1999). Thus, we tested whether theinjection protocols that we used would also label cellscoming from the MGE as they pass through the injectionsite. We performed DiI injections in LGE slices of E14embryos that were subsequently cultured for 48 h. Theslices were selected from the medial level of the rostrocau-dal axis of the brain, where the depth of the interganglionicsulcus is even. In one hemisphere, we placed a DiI crystal inthe LGE (Fig. 5B), while in the other hemisphere, the tracerwas located in the same place (LGE), but after havingremoved the germinative zones of the MGE (Fig. 5C). Thus,the labeled migrating cells could only have come from theLGE, thereby avoiding labeling cells generated in the MGE.When both hemispheres were compared, a similar numberof labeled cells were seen migrating from the LGE throughthe cortical MZ and IZ (Figs. 5B and 5C). However, thepossibility still exists that some migrating cells from theMGE could have been in transit through the LGE when theMGE was removed. To circumvent this possibility, weperformed another experiment ablating the entire MGEbefore setting up the slice culture and inserting the dye 24 hlater. Using this approach, the cells in transit through theLGE had enough time to reach the neuroepithelium beforelabeling the LGE. We thus concluded that the migratorycells that were observed arose in the LGE (Figs. 5D and 5E).Using this approach, we also concluded that cells at themost rostral level of the LGE also contribute to the tangen-tial migratory population, as shown in Fig. 5A.
Tracer injections at different rostrocaudal levels of theganglionic eminences revealed the pathways taken by thecells and the importance of the depth of the interganglionicsulcus in determining whether these cells reach their des-tination (Figs. 2A–2D). Cells generated in the MGE mustcross the LGE to reach the developing neuroepithelium.However, cells generated in the rostral portion of the MGEare confronted with a serious impediment in crossing theLGE due to the depth of the interganglionic sulcus. Conse-quently, BDA injections at rostral levels of the MGE onlylabeled a small population of cells that were struggling toreach to the LGE (Fig. 5F). On the other hand, tracerinjections at the most caudal levels of the MGE labeled alarge number of cells that migrated easily through the LGEto reach the cortical neuroepithelium. These cells weremainly distributed along the PP/MZ and in the lower IZ,although some of them were seen at the VZ (Figs. 5G and 5H).
We also wanted to determine the contribution of eachganglionic eminence to the nascent cerebral cortex. This
FIG. 4. Tracer experiments in rats. (A) DiI injection into the LGE,close to the corticostriatal sulcus, of an E12 embryo cultured for 2days in roller bottle (in toto), showing migrating cells in the PP. (B)DiI implant located in the mediocaudal level of the anteroposterioraxis of the lateral ganglionic eminence in an E14 embryo. Labeledcells appear in the MZ, lower IZ, as well as in the SVZ and VZ. Thisslice was cultured for 2 days. (C) DiI deposits in the mediocaudallevel of the MGE of an E15 embryo. The slice was cultured for 4days. (D) Enlargement of the neuroepithelium in (C). Coronalsections of rat embryos. Dorsal is up, lateral is left. ne, neuroepi-thelium; PP, preplate; v, ventricle. Scale bars: 100 �m in (A, C); 50�m in (B, D).
161Tangential Migration in Neocortical Development
FIG. 5. Tracer experiments. DiI deposits at the rostral level of the rostrocaudal axis (A) and the medial level (B–E) of the LGE in E14 ratbrain slices, were cultured for 2 days. (B, C) The DiI was inserted at the same level of the LGE, but the MGE was removed from the slicepresented in (C) at the time of labeling (black arrow in the inset C). (D) and (E) correspond with another experiment, where the MGE wasablated on the day of the culture (black arrow in the inset D) and the DiI inserted on the following day. (F) BDA injected in the rostral levelof the MGE in an E13 embryo. The injection was performed ex utero, and the embryo was cultured in a roller bottle for 24 h. (G) DiI insertedinto the mediocaudal level of the MGE in an E14 embryo slice, cultured for 3 days. (H) Higher magnification of the neuroepithelium in (G).Coronal sections. Dorsal is up and lateral is left. IZ, intermediate zone; MZ, marginal zone; ne, neuroepithelium; v, ventricle; VZ,ventricular zone. Scale bars, 50 �m.
was clearly seen by injecting DiI into slices of the LGEwhere the VZ and SVZ of the MGE had been removed.These injections were performed at different rostrocaudal
levels of E14 embryos that were then cultured for 2 days. Inthe absence of the MGE, a small number of cells (�) wereseen to migrate through the cortical PP and lower IZ afterDiI injection into the most rostral level of the LGE (Fig. 6A).A few cells trying to reach the cortex were labeled wheninjections were made in the MGE, at rostral levels wherethe interganglionic sulcus is very deep. Similarly, LGEinjections also labeled a few cells (�) in the neuroepithe-lium (Fig. 6B). A large number of cells (���) that hadreached the neuroepithelium were seen when injectionswere performed into the medial level of the MGE, whereboth eminences have attained the same size and the inter-ganglionic sulcus is smooth (Fig. 6C). Injections into theLGE at rostral levels, where the interganglionic sulcus isvery deep, but when the germinative part of the MGE hadbeen removed, labeled a significant number of cells (��)migrating through the cortical MZ and lower IZ (Fig. 6D).Finally, injections at the caudal level, where the intergan-glionic sulcus recedes and two separate ganglionic emi-nences are no longer observed, were performed at twodifferent sites in consecutive slices, one close to the mid-line and the other close to the corticostriatal sulcus. In bothcases, a huge number of tangentially labeled migrating cells(���) were observed throughout the cortical neuroepithe-lium (Figs. 6E and 6F). Thus, we conclude that each gangli-onic eminence contributes to the formation of the cerebralcortex through the generation of different populations oftangential migrating cells. The number of tangential mi-grating cells is larger at caudal levels of the brain, but thereis no significant difference in the contribution of cellsbetween the two eminences.
Trajectories of Cell Migrations Insidethe Neuroepithelium
It is well known that cortical cell generation begins atearly stages of prosencephalic development and consists ofseveral replication waves until the end of embryonic life (inthe rat, E20/E21). However, cells that settle in specificlayers of the neocortex are generated during a short periodof time, in a single round of replication. In utero BDAinjections into the basal telencephalon of an E16 rat embryolabeled migrating cells in the lower IZ at E19 (Figs. 7A–7C).This cell population, coming from the BT, appears from E13and continues to be generated until the end of embryoniclife (and probably also shortly after birth).
Migrating cells have an ovoid-shaped soma with a longleading process that ends in a growth cone, and a fine andshort trailing process. The direction of cell migration can bedetermined from the orientation of the leading process (Fig.7). It is interesting to note that there was considerable cellmovement in opposing directions within the same corticalstratum (PP/MZ). In fact, we identified another populationof cells that were generated in the cortical neuroepitheliumvery early in development. When DiI was implanted insidethe dorsal neuroepithelium of E12 embryos that were
FIG. 6. Schematic representation of the different routes taken bythe cells generated in the ganglionic eminence migrating to thenascent cerebral cortex. Tissue slices were taken at E14 all alongthe rostrocaudal axis of the rat brain. (A) represents the more rostralsection and the drawing in (F) the most caudal section. Selectedsections were labeled with a DiI crystal inserted into the LGE orthe MGE. In (D), the MGE had been removed. Black circlesrepresent the injection sites and the arrows outline the trajectoriesof the different migratory cell populations. The amount of labeledmigrating cells in each experiment is represented with the plussign, � � few cells and ��� � a large quantity of cells.
163Tangential Migration in Neocortical Development
cultured for 24–48 h in roller bottles, cells migratingthrough the PP toward the BT were detected (Figs. 7D and7E). Hemibrains from E13 embryos were DiI-labeled in thedorsal neuroepithelium and cultured for 48 h in collagengel. The image of an embryonic hemibrain viewed from thetop shows cells merging to form a column of tangentiallymigrating cells. These strips of cells are oriented obliquelyin the caudorostral axis, from the injection site to the BT,where the cells are distributed rostrally and caudally, sur-rounding the nascent lateral olfactory tract (Fig. 7F). Parallelexperiments were performed in hemibrains from E14 em-bryos cultured for 48 h, and in all cases, the labeled cellsmigrated tangentially from the injection site downwardthrough the PP, toward the BT, spreading rostrally andcaudally at the level of the olfactory cortex. The course ofthese cell columns is always in a caudorostral direction,independent of the placement of the DiI injection (rostral orcaudal in the anteroposterior axis of the brain).
The expression of GABA, Calbindin, and Calretinin intangentially migrating cells was studied in E13–E17 ratembryos. This period covers the early stages of migrations,although maximal tangential cell migration in the neuro-epithelium was observed at E16 and E17.
GABA. Expression of GABA began in migrating cells atE14 (Figs. 8A and 8E). At this age, the mass of tangentialmigratory cells that arrived at the cortical IZ adopted acuneiform shape (similar to that adopted by the CP) as itadvanced through the IZ of the cortical neuroepitheliumfollowing a lateromedial gradient. At the same time,GABAergic cells advance tangentially through the PP. AtE15 (Figs. 8B and 8F), these cells adopted a more dorsalposition, although always following the growth and ad-vancement of the CP cells. The GABAergic populationreaches the most dorsal part of the neuroepithelium at E16
(Figs. 8C and 8G). Some of these cells settle in the lower IZ,directing their leading processes to either the PP (Fig. 8F) orthe SP (Fig. 8G), according to the developmental stage. Thismight suggest that these cells continue their migrationthrough the neuroepithelial thickness to reach their finalposition in different layers of the neocortex. However, fewcells in the lower IZ had their leading processes directedtoward the VZ, as if they were descending (Figs. 8F and 8G).At this age, cells in two other regions also express GABA:the SP and the MZ. At E17 (Figs. 8D and 8H), the cuneiformdisposition adopted by the GABA-positive cells had van-ished, as they occupied the whole lateromedial extension ofthe neuroepithelium, including the primordium of thehippocampus. The horizontal disposition of the GABAergiccells in the developing hippocampus, and the connectionwith the tangential migrating cells, suggests that thisstructure could be another target for cells generated in theganglionic eminence. In addition, at E15, E16, and E17,some GABAergic cells could be seen in the neuroepithelialVZ and SVZ with the classic morphology of tangentialmigratory cells.
Calbindin. At E14, when the first GABAergic cells werelocated in the neuroepithelium, there were no Calbindin(CB)-immunoreactive cells in the IZ, although there weremany fibers that contained CB running through this stra-tum. At E15, CB-expressing fibers were still seen runningthrough the IZ, as well as some immunolabeled cells closeto the corticostriatal sulcus. In contrast, at E16 (Fig. 8I) alarge number of CB-containing cells populate the cerebralcortex from the corticostriatal sulcus to the most dorsalpart of this structure, displaying the typical horizontaldisposition of tangential migrating cells. Later on, CB-labeled cells were observed in the MZ, CP, SP, and thelower IZ. Some immunolabeled cells were detected in theVZ and SVZ, showing the tangential migratory morphol-ogy, as occurred with GABAergic cells. At E17 (Fig. 8J), theCB-containing cells remained in the above mentioned ar-eas, but cells migrating through the IZ were dispersed in
FIG. 7. (A–C) E19 rat. In utero BDA injection into the basal telencephalon at E16 identifies cells migrating through the lower intermediatezone (LIZ). Gray arrows point to the course of migration, also marked by the leading process. (D–F) DiI deposits in the dorsalneuroepithelium (ne). (D) E12 embryo cultured in roller bottles for 24–48 h. Coronal section was counterstained with bisbenzimide. Highermagnification of the labeled cells in (E) shows the cells running through the preplate (PP) towards the basal telencephalon. (F) DiI in thedorsal neuroepithelium (white arrowhead in the inset) of an E13 hemibrain cultured for 48 h in collagen. Dorsal is up and lateral is left in(A–E). Rostral is to the right and dorsal is to the top in (F). LIZ, lower intermediate zone; ne, neuroepithelium; PP, preplate, SVZ,subventricular zone; v, ventricle; VZ, ventricular zone. Scale bars: 100 �m in (D, F); 250 �m in (A, B, E); 10 �m in (C).FIG. 8. Cell markers. GABA expression in coronal (A–C, E–G) and sagittal (D, H) sections of rat embryos. (A) E14 embryo showingGABAergic cells running through the PP and the IZ. (E) Enlargement of the box in (A). (B, F) E15 embryo. GABAergic cells colonized thePP and IZ. Box in B is enlarged in F. C, G: E16 embryo. The cortical plate splits the PP leading to the appearance of the MZ and SP. (G)Enlargement of the box in (C). (D, H) E17 rat embryo showing GABAergic hippocampal cells in the ganglionic tangential migrationpathway. (H) shows the enlargement of the box in (D). (I, J) E16 embryo. (I) Calbindin-positive cells occupy the MZ, SP, and LIZ. At E17(J), Calbindin-positive cells are dispersed in the entire IZ. Coronal sections. (K, L) Coronal sections of rat embryos stained with theCalretinin antibody. At E16 (K), there are Calretinin-immunoreactive cells in the MZ and SP, but none are present in the IZ. At E17 (L),some Calretinin-positive cells appear in the LIZ.
165Tangential Migration in Neocortical Development
this stratum. It is interesting to note that CB-immuno-reactive cells reach the hippocampal primordium, as de-scribed for GABAergic cells.
Calretinin. Corticosubcortical projecting axons beginto cross the corticostriatal sulcus at E14 in the rat (E12.5 inthe mouse). These axons, coming from the PP layer, containCalretinin, and settled in the same cortical area as tangen-tial migratory cells at E14. At this age, there wereCalretinin-labeled cells surrounding the lateral olfactorytract, with a horizontal disposition and with their leadingprocesses pointing down toward the ventral telencephalon.In addition, the expression of Calretinin in the PP is veryintense throughout the whole neuroepithelium. Some dayslater, at E16, Calretinin labels projecting cells in the CP andSP (Fig. 8K), although a few horizontal migrating cells wereseen dispersed through the IZ of E17 rats (Fig. 8L).
In summary, GABA, Calbindin, and Calretinin are ex-pressed in tangential migrating cells at different times andin a sequential order, suggesting the existence of differentcell populations with different roles in cortical develop-ment.
DISCUSSION
In this study, we have introduced the concept that theanteroposterior extension of the ganglionic eminences andthe interganglionic sulcus are extremely important in regu-lating tangential migration to the cortex of developing ratembryos. We have demonstrated that both ganglionic emi-nences (LGE and MGE) give rise to a substantial number ofcells that migrate tangentially, following several routes toreach different layers of the mature cerebral cortex. Inaddition, an opposing migratory pathway is described forcells generated specifically in the telencephalic vesicles,which cross the corticostriatal sulcus to enter the BT andcolonize the piriform cortex. Cells that migrate tangentiallyfrom the BT express molecules such as GABA, Calbindin,and Calretinin in a sequential fashion.
Ganglionic Eminences and the Generationof Tangential Migrating Cells
The first macroscopic sign of an eminence, the MGE, canbe seen in the rat at E12, coinciding with the expression ofthe homeobox gene Nkx2.1 in the mouse (Sussel et al.,1999). The LGE emerges at E12.5 (present observations),and this short time-lapse between the appearance of the twoeminences may not be relevant for cell generation andmigration. Indeed, the MGE expands until E14, at whichpoint it begin to decrease in size until E17, when the BT isformed. In contrast, the LGE grows progressively until theMGE fades away. It is important to bear in mind that theLGE extends along the whole rostrocaudal axis, while theMGE emerges rostrally in the second third of the axis.Moreover, the interganglionic sulcus is very deep at rostral
levels, separating both the eminences and making it diffi-cult to follow the course of the cells that originate in theMGE as they pass through the LGE toward the corticalneuroepithelium. For this reason, there is a large number ofmigrating cells that come from the medial to caudal levelsof the rostrocaudal axis.
Angevine and Sidman (1961) described the existence ofcortical cells generated in the cortical neuroepitheliumwith an inside-out gradient. However, Cajal-Retzius cellsdifferentiate before the formation of the cortical plate andpopulate the most external layer (Cajal, 1890). Marın-Padilla (1971, 1972, 1978) emphasized the early differentia-tion of Cajal-Retzius and subplate cells, as was later con-firmed by autoradiography (Konig et al., 1977; Rickmann etal., 1977; Caviness, 1982). Cortical-generated cells migrateradially to their cortical stratum, with their cellular bodiesand processes being oriented along the vertical axis (Ange-vine and Sidman, 1961). However, some cells migratetangentially, adopting a horizontal disposition in the differ-ent strata (Hataı, 1902), predominantly in the lower IZ(Stensaas, 1967). Different authors assumed that these cellsoriginated in the cortical VZ and SVZ (Van Eden et al.,1989; Bayer and Altman, 1991; Fishell et al., 1993; De Diegoet al., 1994). Nevertheless, the latest findings indicate thatsome of the cells generated in the ganglionic eminencecross the corticostriatal sulcus to become part of the cere-bral mantle (De Carlos et al., 1996; Anderson et al., 1997;Tamamaki et al., 1997). Most cortical cells are generated inthe VZ and SVZ of the cortical neuroepithelium, and somemitotic structures are present in or very near the PP,suggesting that some MZ neurons are generated locally(Valverde et al., 1995). Additionally, many cortical cellscome from the BT. The incorporation of BrdU indicates thattangentially migrating cells are located along the PP be-tween E13 and E14, and in the lower IZ during a moreextended period of time (beginning at E13 with generationreaching a peak at E15–E16). These data are importantbecause the cortical cells that migrate radially are generatedsequentially in the VZ and SVZ of the neuroepitheliumduring short periods. Cells destined to occupy differentcortical layers are generated with an inside-out gradient, theyounger cells occupying the upper layers. However, thegeneration of tangential migratory cells begins at E13 andcontinues throughout embryonic development and prob-ably shortly after birth. This is in agreement with thetiming of rat striatal cell generation, reported to occurbetween E12 and postnatal day 2, peaking around E15(Marchand and Lajoie, 1986).
The fact that cells generated in the ganglionic eminenceare incorporated into the cortical mantle during the forma-tion of the different cortical layers could have severalimportant implications. First of all, we must consider therole that these cells play throughout the entire period ofcortical development. Second, most are GABAergic cellsthat integrate into the developing cortex when pyramidal
cells are occupying the different cortical layers while pro-jecting toward their targets. This inhibitory cell invasion isprobably necessary for pyramidal cell maturation. Third,the origin of these cells at different times and from differentcomponents of the BT implies a heterogeneous characterand different roles.
Characterization of Different Populationsof Tangential Migrating Cells
We show that at least two different cortical cell popula-tions are generated in each eminence of the BT that willsubsequently migrate through the PP/MZ and lower IZ. Asin the cortical neuroepithelium, the cells originate in theVZ and continue to be generated in the SVZ. Each cellpopulation arises at a different time, from E13 to earlypostnatal (Marchand and Lajoie, 1986). Several groups havealready described similar types of migration and proposedthat GABAergic-cortical cells are generated in the gangli-onic eminences (Brunstrom et al., 1997; Anderson et al.,1997a, 2001; Tamamaki et al., 1997). Moreover, it has beenshown that the homeobox-containing transcription factorsDlx1 and Dlx2 are needed for their migration (Anderson etal., 1997a,b, 1999, 2001; Marın et al., 2000). We hadproposed that cells generated in the LGE cross the cortico-striatal sulcus to form part of the cortical mantle (De Carloset al., 1996). However, it has been suggested by otherauthors that the origin of these cells was the MGE (Lavdaset al., 1999; Wichterle et al., 1999) and that these cells hadtaken up the dye as they passed through the LGE injectionsite. However, we strongly believe that the origin of thesecells is the LGE for several reasons. First, our experimentswere undertaken in embryos in toto, rather than in tissueslices whose thickness probably does not allow a free andcomplete cell migration. Second, we performed the injec-tions at very rostral levels, where the MGE either does notexist, or is very far from the LGE, rather than injecting atmore caudal levels where the eminences come together.Neither site was precise, because the depth of the sulcus atrostral levels could impair the migration from the MGEthrough the LGE. Nevertheless, we thought it was neces-sary to study the anatomical appearance of both eminencesand the rostrocaudal progression of the interganglionicsulcus, and found that the extension of the ganglioniceminences along the rostrocaudal axis and the depth of thesulcus influence the tangential migration. We believe thatthese data will be useful to other scientists in the field.Another interesting finding of this study is that some earlygenerated cells in the VZ/SVZ of the medial to dorsalcortical neuroepithelium use the PP to migrate tangentiallydownward toward the BT, following a rather caudorostralcourse. The final destination of these cells seems to be theolfactory cortex, surrounding the lateral olfactory tract.This population of cells ( Jimenez et al., 2000), might be thesame population of cells reported by Tomioka et al. (2000),and their role might be to form the cell-free channel used by
growing axons forming the lateral olfactory tract (Lopez-Mascaraque et al., 1996).
At E14, GABAergic cells are present in the PP and IZ, andsoon after, the IZ GABA-population is seen running alongwith the nascent CP. This convergence continues until E16,when both populations reach the most dorsal part of thetelencephalon, suggesting an interaction between the two.Some IZ migrating cells seem to enter into the CP as theyadvance, indicating that the tangential migrating cells arenot transitory, as reported elsewhere (Tamamaki et al.,1997). Furthermore, the integration of the tangentialGABAergic cells into the CP has been described previously,before the precise origin of tangential migrating cells wasknown, and when the number of GABAergic cells in the CPwas seen to increase simultaneous to a decline in theIZ/SVZ at E19 (Van Eden et al., 1989). However, a smallnumber of cells point their leading processes toward the VZ(Nadarajan et al., 2000). At E17, GABAergic cells becomevisible in the hippocampal anlage, as a continuation of thetangential migratory cells. This suggests that this structureis also a target for some cells generated in the BT (Pleasureet al., 2000). We described the sequential expression ofGABA, Calbindin, and Calretinin in the cortex, althoughLavdas et al. (1999) previously denied the existence in thecortex of Calretinin-immunoreactive tangential migratingcells. This sequential expression, together with the differ-ent timing of generation in the distinct BT structures (MGEand LGE), leads us to propose the existence of severalpopulations of migrating cells, rather than one single popu-lation, that express different markers at different times. Infact, the three markers tested coexist in the neuroepithe-lium during several embryonic days, from E17.
In summary, we have addressed the macroscopic devel-opment of the ganglionic eminences and have shown theimportance of the interganglionic sulcus in the tangentialmigration of cells from the eminences to the cortex. Weconfirm that cortical cells are generated in the MGE anddemonstrate their early generation in the LGE. In this way,both eminences give rise to a substantial population ofcortical cells that merge and constitute a heterogeneous cellpopulation, bearing different markers and probably execut-ing different roles in the early development of the neocor-tex. Among the role of these cell populations, we suggest afunctional interaction between the developing cortical plateand the GABAergic tangential migratory cells that populatethe intermediate zone. This interaction may be critical for thegrowth and development of the cortical plate.
ACKNOWLEDGMENTS
We thank A. Nieto for helpful comments on the manuscript, M.Sefton for editorial assistance, and M. L. Poves for technicalsupport. Grant sponsors: Ministerio de Educacion y Cultura (MEC)of Spain: PB96-0813 and Consejerıa de Educacion y Cultura de laComunidad de Madrid: 08.5/0037/1998. D.J. is a predoctoral fellowfrom MEC.
167Tangential Migration in Neocortical Development
Anderson, S. A., Eisenstat, D. D., Shi. L., and Rubenstein, J. L. R.(1997a). Interneuron migration from basal forebrain to neocortex:Dependence on Dlx genes. Science 278, 474–476.
Anderson, S. A., Qiu, M-S., Bulfone, A., Eisenstat, D. D., Meneses,J., Pedersen, R., and Rubenstein, J. L. R. (1997b). Mutation of thehomeobox genes Dlx-1 and Dlx-2 disrupt the striatal subven-tricular zone and differentiation of late-born striatal neurons.Neuron 19, 27–37.
Anderson, S. A., Mione, M., Yun, K., and Rubenstein, J. L. R. (1999).Differential origins of neocortical projection and local circuitneurons: Role of Dlx genes in neocortical interneurogenesis.Cereb. Cortex 9, 646–654.
Anderson, S. A., Marın, O., Horn, C., Jennings, K., and Rubenstein,J. L. R. (2001). Distinct cortical migrations from the medial andlateral ganglionic eminences. Development 128, 353–363.
Angevine, J. B., and Sidman, R. L. (1961). Autoradiographic study ofcell migration during histogenesis of cerebral cortex in themouse. Nature 192, 766–768.
Bayer, S. A., and Altman, J. (1991). “Neocortical Development.”Raven Press, New York.
Brunstrom, J. E., Gray-Swain, R. M., Osborne, P. A., and Pearlman,A. L. (1997). Neuronal heterotopias in the developing cerebralcortex produced by neurotrophin-4. Neuron 4, 505–517.
Cajal, S. R. (1890). Textura de las circunvoluciones cerebrales de losmamıferos inferiores. Nota preventiva. Gaceta Medica Cata-lana, 15 de Diciembre, 22–31.
Caviness, V. S. (1982). Neocortical histogenesis in normal andreeler mice: A developmental study based upon 3H-thymidineautoradiography. Dev. Brain Res. 4, 293–302.
Chedotal, A., Del Rıo, J. A., Ruız, M., He, Z., Borrell, V., de Castro,F., Ezan, F., Goodman, C. S., Tessier-Lavigne, M., Sotelo, C., andSoriano, E. (1998). Semaphorins III and IV repel hippocampalaxons via two distinct receptors. Development 125, 4313–4323.
Cockroft, D. L. (1990). Dissection and culture of postimplantationembryos. In “Postimplantation Mammalian Embryos. A Practi-cal Approach” (A. J. Copp and D. L. Cockroft, Eds.), pp. 15–40.Oxford Univ. Press, New York.
De Carlos, J. A., and O’Leary, D. D. M. (1992). Growth and targetingof subplate axons and establishment of major cortical pathways.J. Neurosci. 12, 1194–1211.
De Carlos, J. A., Lopez-Mascaraque, L., and Valverde, F. (1996).Dynamics of cell migration from the lateral ganglionic eminencein the rat. J. Neurosci. 16, 6146–6156.
De Diego, I., Smith-Fernandez, A., and Fairen, A. (1994). Corticalcells that migrate beyond area boundaries: Characterization of anearly neuronal population in the lower intermediate zone ofprenatal rats. Eur. J. Neurosci. 6, 983–9974.
Fishell, G., Mason, C. A., and Hatten, M. E. (1993). Dispersion ofneural progenitors within the germinal zones of the forebrain.Nature 362, 636–638.
Fishell, G. (1995). Striatal precursors adopt cortical identities inresponse to local cues. Development 121, 803–812.
Hataı, S. (1902). Observations on the developing neurons of thecerebral cortex of foetal cats. J. Comp. Neurol. 12, 199–204.
Honig, M. G., and Hume, R. I. (1986). Fluorescent carbocyaninedyes allow living neurons of identified origin to be studied inlong-term cultures. J. Cell Biol. 193, 171–187.
Honig, M. G., and Hume, R. I. (1989). Carbocyanine dyes. Novelmarkers for labelling neurons. Trends Neurosci. 12, 336–338.
Jimenez, D., Lopez-Mascaraque, L., Valverde, F., and De Carlos,J. A. (1999a). El telencefalo basal: Fuente de neuronas para lacorteza cerebral. Revista de Neurologıa 30, 216.
Jimenez, D., Lopez-Mascaraque, L., Valverde, F., and De Carlos,J. A. (1999b). New analysis on migratory routes from the basaltelencephalon. Soc. Neurosci. Abstr. 25, 2035.
Jimenez, D., Lopez-Mascaraque, L., Valverde, F., and De Carlos,J. A. (2000). Tangential migrations in neocortical development.Soc. Neurosci. Abstr. 26, 570.
Konig, N., Valat, J., Fulcrand, J., and Marty, R. (1977). The time oforigin of Cajal-Retzius cells in the rat temporal cortex. Anautoradiographic study. Neurosci. Lett. 14, 21–26.
Lavdas, A. A., Grigoriou, M., Pachnis, V., and Parnavelas, J. G.(1999). The medial ganglionic eminence gives rise to a populationof early neurons in the developing cerebral cortex. J. Neurosci.15, 7881–7888.
Lopez-Mascaraque, L., De Carlos, J. A., and Valverde, F. (1996).Early onset of the rat olfactory bulb projections. Neuroscience 70,255–266.
Marchand, R., and Lajoie, L. (1986). Histogenesis of the striopallidalsystem in the rat. Neurogenesis of its neurons. Neuroscience 17,573–590.
Marın, O., Anderson, S. A., and Rubenstein, J. L. R. (2000). Originand molecular specification of striatal interneurons. J. Neurosci.20, 6063–6076.
Marın-Padilla, M. (1971). Early prenatal ontogenesis of the cerebralcortex (neocortex) of the cat (Felix domestica). A Golgi study. I.The primordial neocortical organization. Z. Anat. Entw. 134,117–145.
Marın-Padilla, M. (1972). Prenatal ontogenetic study of the princi-pal neurons of the neocortex of the cat (Felix domestica). A Golgistudy. II. Developmental differences and their significances. Z.Anat. Entw. 136, 125–142.
Marın-Padilla, M. (1978). Dual origin of the mammalian neocortexand evolution of the cortical plate. Anat. Embryol. 153, 109–126.
Nadarajan, B., Wong, R. O. L., and Parnavelas, J. G. (2000). Neuro-nal migration towards the ventricular zone of the developingneocortex. Soc. Neurosci. Abstr. 26, 569.
Nguyen Ba-Charvet, K. T., Brose, K., Marillat, V., Kidd, T., Good-man, C. S., Tessier-Lavigne, M., Sotelo, C., and Chedotal, A.(1999). Slit2-mediated chemorepulsion and collapse of develop-ing forebrain axons. Neuron 22, 463–473.
Pleasure, S. J., Anderson, S., Hevner, R., Bagri, A., Marın, O.,Lowenstein, D. H., and Rubenstein, J. R. (2000). Cell migrationfrom the ganglionic eminences is required for the development ofhippocampal GABAergic interneurons. Neuron 28, 727–740.
Rakic, P. (1971). Guidance of neurons migrating to the fetalmonkey neocortex. Brain Res. 3, 471–476.
Rakic, P. (1972). Mode of cell migration to the superficial layers offetal monkey neocortex. J. Comp. Neurol. 145, 61–84.
Rickmann, M., Chronwall, B. M., and Wolff, J. R. (1977). On thedevelopment of non-pyramidal neurons and axons outside thecortical plate: The early marginal zone as a pallial anlage. Anat.Embryol. 151, 285–307.
Schutte, B., Reynders, M. M., Bosman, F. T., and Blijham, G. H.(1987). Effect of tissue fixation on anti-bromodeoxyuridine im-munohistochemistry. J. Histochem. Cytochem. 35, 1343–1345.
Smart, I. H. M., and Sturrock, R. R. (1979). Ontogeny of theneostriatum. In “The Neostriatum” (I. Divac and R. G. E. Oberg,Eds.), pp. 127–146. Pergamon Press, Oxford.
Stensaas, L. J. (1967). The development of hippocampal and dorso-lateral pallial regions of the cerebral hemisphere in fetal rabbits.II. Twenty millimeter stage, neuroblast morphology. J. Comp.Neurol. 129, 71–84.
Stewart, G. R., and Pearlman, A. L. (1987). Fibronectin-like immu-noreactivity in the developing cerebral cortex. J. Neurosci. 7,3325–3333.
Sussel. L., Marın, O., Kimura, S., and Rubenstein, J. R. (1999). Lossof Nkx2.1 homeobox gene function in a ventral to dorsal mo-lecular respecification within the basal telencephalon: Evidencefor a transformation of the pallidum into the striatum. Develop-ment 126, 3359–3370.
Tamamaki, N., Fujimori, E., and Takauji, R. (1997). Origin androute of tangentially migrating neurons in the developing neo-cortical intermediate zone. J. Neurosci. 17, 8313–8323.
Tomioka, N., Osumi, N., Sato, Y., Inoue, T., Nakamura, S.,Fujisawa, H., and Hirata, T. (2000). Neocortical origin and
tangential migration of guidepost neurons in the lateral olfactorytract. J. Neurosci. 20, 5802–5812.
Van Eden, C. G., Mrzljak, L., Voorn, P., and Uylings, H. B. M.(1989). Prenatal development of GABA-ergic neurons in theneocortex of the rat. J. Comp. Neurol. 289, 213–227.
Valverde, F., Lopez-Mascaraque, L., Santacana, M., and De Carlos,J. A. (1995). Persistence of early-generated neurons in the rodentsubplate: Assessment of cell death in neocortex during the earlypostnatal period. J. Neurosci. 15, 5014–5024.
Wichterle, H., Garcıa-Verdugo, J. M., Herrera, D. G., and Alvarez-Buylla, A. (1999). Young neurons from medial ganglionic emi-nence disperse in adult and embryonic brain. Nat. Neurosci. 2,461–466.
Received for publication October 31, 2001Revised December 17, 2001
Accepted December 17, 2001Published online February 25, 2002
169Tangential Migration in Neocortical Development
