Recent studies have shown that the two main types of cortical neurons, pyramidal and nonpyramidal, have different origins and use different migratory routes — radial and tangential respectively. The role of neurotransmitters in radial migration is well known; however, there are no data about their effect on intracellular calcium — [Ca2+]i — in tangentially migrating cells. We have performed ratiometric and confocal calcium imaging of 1,1′-dioctodecyl-3,3,3′,3′-tetramethylindocarbocyanine labelled tangentially migrating neurons in the intermediate zone cells of fetal rat coronal slices. Superfusion with N-methyl-d-aspartic acid (NMDA) leads to an increase in [Ca2+]i, which is blocked by the antagonist APV or the presence of Mg2+ in the medium. Kainate produced an increase in [Ca2+]i that could be blocked by the non-NMDA antagonist CNQX. Muscimol, an agonist of GABAa-receptors, produced a transitory increase in [Ca2+]i that was blocked by the specific antagonist bicuculline or the presence of tetrodotoxin in the medium. We conclude that tangentially migrating cells display consistent [Ca2+]i changes in response to agonists of NMDA, non-NMDA and GABAa receptors, suggesting that these cells are quite mature and homogeneous. The endogenous activation of these receptors may have either a direct effect on tangential migration or modulate the response of migrating cells to external cues.
Formation of the six layered mammalian cortex is the result of a complex process of neuronal generation, migration and establishment of appropriate synapses (McConnell, 1995). Recent evidence indicates that different neuronal phenotypes — pyramidal and nonpyramidal — originate in distinct proliferative zones and migrate along different paths (Pearlman et al., 1998; Anderson et al., 1999; Parnavelas, 2000). The neuroblasts destined to be projection pyramidal neurons are born in the ventricular zone (VZ) and follow a radial migration along the glia scaffolding (Hatten, 1999). From the proliferative neuro-epithelium, these immature neurons migrate towards the cortical surface (Nadarajah et al., 2001), forming the cortical plate (CP) where they are deposited according to an inside-out gradient (Angevine and Sidman, 1961; Bayer and Altman, 1990). By contrast, inhibitory GABAergic interneurons follow tangential migratory paths in which the cells run perpendicular to radial glial processes. Cells migrate from their birthplace — the ganglionic eminence of the basal telencephalon — towards the developing cortex were they settle at different levels (de Carlos et al., 1996; Anderson et al., 1997, 2001; Tamamaki et al., 1997; Lavdas et al., 1999; Wichterle et al., 1999, 2001).
Migrating cells respond to short and long range signals in order to reach appropriate territories and establish the right connections. Although the leading tip must respond to different cues in radial and tangential migration (Gleeson and Walsh, 2000; Walsh and Goffinet, 2000), the same cytoplasmic mechanisms for detecting and transducing guidance signals may operate in both kinds of migration (Song and Poo, 2001). Axon extension and pathfinding is regulated by diverse environmental cues (Tessier-Lavigne and Goodman, 1996) that act through common transduction pathways in which calcium signalling plays a central role (Gomez et al., 1995; Komuro and Rakic, 1996; Goldberg and Grabham, 1999; Gomez and Spitzer, 1999).
Neurotransmitters are known to play an important role in radial migration (Rakic and Komuro, 1995). Thus, glutamate acting at N-methyl-d-aspartic acid (NMDA) receptors (Behar et al., 1999) stimulates neuronal chemotaxis (migration along a chemical gradient) and NMDA antagonists decrease labelled cells in the outer intermediate zone (Hirai et al. 1999), indicating that blockade of NMDA receptors inhibits neuronal migration. Gamma-aminobutyric acid (GABA) stimulates chemotaxis and chemokinesis (increased random movement) in different cell populations (Behar et al., 1996, 1998). It is well established that the effects of neurotransmitters on neuronal migration are mediated by changes in intracellular calcium concentration (Rakic and Komuro, 1995; Behar et al., 1996, 1998; Komuro and Rakic, 1996, 1998).
There is little information concerning the physiology of cells migrating tangentially. Recently, it has been shown (Metin et al., 2000) that GABAergic calbindin-positive intermediate zone (IZ) cells express inwardly rectifying calcium-permeable AMPA receptors, but not NMDA receptors. Likewise, these cells display electrophysiological responses to GABAa agonists. In the same study, it was shown that corticofugal growth cones are tightly apposed to calbindin-positive IZ cells, suggesting that glutamate release may activate AMPA receptors, promote a rise in [Ca2+]i and regulate tangential migration. Furthermore, activation of AMPA receptors leads to neurite retraction of microtubule associated protein (MAP-2) positive cells (Poluch et al., 2001), suggesting a possible role for these receptors in tangential migration. However, no study has yet documented the effects of glutamate or other neurotransmitters on [Ca2+]i signalling in tangentially migrating cells.
The experiments presented here investigate the effects of activation of glutamate and GABA receptors on [Ca2+]i in identified tangentially migrating cells as a first step in the study of the relationship between calcium dynamics and tangential migration. We have found that tangentially migrating cells in the IZ display consistent [Ca2+]i changes when challenged with agonists of glutamate and GABA receptors.
Materials and Methods
Wistar albino rats were mated overnight and vaginal smears examined the next morning. The day of sperm positivity was taken as embryonic day 0 (E0). Pregnant rats were deeply anaesthetized with chloral hydrate (i.p.) and the fetuses extracted by Caesarean section. Experimental procedures involving live animals were carried out in accordance with the guidelines set by the European Community and were approved by the Animal Care Committee of the authors’ institution.
Fetuses (E16) were decapitated and heads immediately placed in chilled artificial cerebrospinal fluid (ACSF) — containing (in mM): NaCl, 124; KCl, 5; KH2PO4, 1.2; MgSO4, 1.3; CaCl2, 2.4; and glucose, 10 — bubbled with 95% O2 and 5% CO2 to reach a final pH of 7.3. Whole brains were dissected out and embedded in warm (41°C) 4% low-melting-point agarose (Sigma, St Louis, MO) and rapidly cooled. Coronal cortical slices 300 μm thick were cut with a Vibratome (Leica VT1000S, Germany). Slices were placed for 1 h in 1 ml of DM EM/F12 (Sigma) culture medium with 6.5 mg/ml glucose, 0.1 mM glutamine (Sigma), 50 mg/ml penicillin, 50 mg/ml streptomycin and 10% fetal calf serum (FCS; Life Technologies, Gaithersburg, MD) on top of Millicell membranes (Millipore, Bedford, MA) of 0.4 μm pore diameter. After this initial period, the cultures were kept in Neurobasal medium (Life Technologies) supplemented with B27 (1:50; Life Technologies), 6.5 mg/ml glucose, 0.1 mM glutamine and 50 mg/ml penicillin/streptomycin.
Caudal sections were used in which both the medial ganglionic eminence (MGE) and lateral ganglionic eminence (LGE) were present. To identify tangentially migrating cells, small crystals of 1,1′-dioctodecyl-3,3,3′,3′-tetramethylindocarbocyanine (DiI; Molecular Probes, Eugene, OR) dissolved in ethanol and dried were placed, with the help of a dissecting scope, in either the MGE or LGE in a position close to the corticostriatal sulcus. The slices were cultured for periods ranging from 24 to 48 h. For calcium measurements, we chose labelled cells in the IZ that were isolated. For the experiments described in this study, 75 slices from 45 animals were processed.
We used either ratiometric Fura-2 [Ca2+]i measurements, that allowed a precise quantification of response, or confocal Fluo-3 microfluorescence for improved spatial resolution. Experiments were carried out on slices at 23–26°C continuously superfused (1 ml/min) in ACSF medium. No qualitative differences were observed in response to agonists at E16 or E16 + 48 h. Therefore, data presented in the results have been pooled. Drugs were applied through a pumped line at a rate of 4 ml/min, during the times indicated in the figures.
Ratiometric Fura-2 Measurements
Cortical slices were removed from the culture medium and incubated for 1 h in Neurobasal medium containing 10 μM Fura-2 AM (Molecular Probes, Eugene, OR), dissolved in 0.09% dimethysulfoxide (DMSO) and 0.006% pluronic acid as previously described (Martinez-Galan et al., 2001). Slices were transferred to the stage of an upright Zeiss Axioskop microscope and viewed through 60× or 20× water immersion objectives. Slices were excited at 340 and 380 nm and the fluorescence emitted at 510 nm recorded by a Hamamatsu C2400 intensifier–Dage 72 video camera. Time-lapse images were acquired every 5 s and spatially analyzed using a MCID M4 System (Imaging Research Inc., St Catherine’s, Ontario, Canada). No background subtraction or filtering was applied to the images. [Ca2+]i was calculated by interpolation of the ratio images into a look-up table constructed by imaging Fura-2 solutions with known calcium concentration in the same experimental set-up, using the following equation (Grynkiewicz et al., 1985):
Confocal Fluo-3 Measurements
Cortical slices were removed from the culture medium and incubated for 1 h in Neurobasal medium containing 10 μM Fluo-3 AM (Molecular Probes, Eugene, OR), dissolved in 0.09% DMSO and 0.006% pluronic acid. Slices were transferred to the stage of an upright Leica DMLFSA microscope coupled to a confocal spectral scanning head (Leica TCS SL) and viewed through 60×, 40×, 20× or 10× water immersion objectives. DiI labelling was excited with the 543 nm line of a neon laser and the fluorescence emitted between 555 and 700 nm measured by one of the photomultipliers (PMT). Fluo-3 labelling was excited with the 488 line of an argon laser and the fluorescence emitted between 500 and 540 nm measured by another PMT. These settings were chosen in control experiments to minimize cross-talk between channels; therefore, the contribution of the DiI fluorescence to Fluo-3 fluorescence is negligible in our experiments. Direct images were obtained by a third PMT to situate the analyzed region of the slice. Changes in Fluo-3 cell fluorescence were expressed as the ratio between fluorescence at the beginning of the experiment and at a given time-point.
Fluo-3 images were acquired with maximum confocality according to the objective magnification used. Time-lapse images were acquired every 3 s and the changes in fluorescence measured in the soma of labelled cells. After recording the response to agonist(s) we performed an xyz scan in order to have the complete morphology of the cell under study. In most of the examples, the reconstructed cell expanded between 20 and 40 μm in the z-axis. In this study, we aimed to image isolated DiI labelled cells. To avoid photodamage of the cells and photobleaching of the dyes, the laser intensity was adjusted to the minimum compatible with a good signal-to-noise ratio.
NMDA, (s)-α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), (2s-(2α,3β,4β))-2-carboxy-4-(1-methylethenyl)-3-pyrrolidineacid acid (kainate), 5-aminomethyl-3-hydroxyisoxazole (muscimol), (±)-2-amino-5-phosphonovaleric acid (APV), 6-cyano-7-nitroquinoxaline-2, 3-dione (CNQX), (R-(R*,S*))-6-(5,6,7,8,-tetrahydro-6-methyl-1,3-dioxolo-(4,5-g) isoquinolin-5-yl) furo (3,4-e)-1,3-benzodioxol-8(6H)-one (bicuculine), tetrodotoxin (TTX) were purchased from Tocris (Essex, UK). Drugs were prepared as stock solutions in distilled H2O or 1 N NaOH according to manufacturers’ indications. All drugs were stored at 4°C and prepared to working concentrations in ACSF daily.
Numerical data were statistically analyzed using one-way analysis of variance (ANOVA).
The IZ, located between the cortical plate (CP) and the subventricular zone (SVZ), consists of tangentially oriented fibres and a diverse population of cells that include those migrating radially towards the CP and others migrating tangentially from extracortical origins. In the first set of experiments we have characterized the response to neurotransmitters of identified tangentially migrating cells in the IZ. Subsequently, we have characterized the pharmacological profile of the responses in IZ cells.
Calcium Responses in Tangentially Migrating Cells
To identify tangentially migrating cells, small crystals of DiI were placed in the GE and the slices cultivated for periods of time ranging from 24 to 48 h. The initial experiments were done using ratiometric Fura-2 loaded slices. These experiments were complemented with confocal microscopy and gave qualitatively the same results. Due to the fact that confocal microscopy renders a much superior spatial resolution, only the results obtained with this approach are presented.
Most of the DiI labelled cells in the IZ (10 out of 13) were bipolar with tangentially directed processes (see below). Their morphology is very similar to that found in previous inmuno-cytochemical studies on tangentially migrating cells (O’Rourke et al., 1995; Tamamaki et al., 1997; Lavdas et al., 1999). The cells possessed a long process directed dorsomedially which in some cases was branched. On the opposite side of the soma, the process was shorter with a trailing appearance. The remaining three cells had a radial disposition with processes towards the CP and VZ. In these cells, a trailing process was also visible. The results presented in the following sections apply to both cell types.
Figure 1 shows the effect of NMDA on [Ca2+]i in a tangentially migrating DiI labelled cell in the IZ. Figure 1A shows a low-power magnification of the slice. The DiI was deposited in the ganglionic eminence, indicated by an arrow, and the cells allowed to migrate for 24 h. The DiI labelled stream of cells is represented in red and the framed area of the photomicrograph corresponds to the zone analyzed in the subsequent panels. Figure 1B shows the DiI fluorescence image of a tangentially migrating cell. The cell has a migratory appearance with a bifurcated leading process pointing dorsomedially. There is also a trailing process at the opposite pole of the soma. Figure 1C shows the colocalized fluorescence of the calcium indicator Fluo-3 (green) and DiI (red). As expected, in addition to the DiI labelled cell there are many other cells loaded with the calcium indicator. The record displayed in Figure 1D shows the time-course of the labelled cells calcium response to NMDA. This record was obtained using the theoretical settings of maximum confocality, in a single plane of best cell somata focusing, with no integration in the z-axis. NMDA was applied to the bathing medium (containing no Mg2+; see below) during the time indicated by the bar. The increase in fluorescence reflects an increase in [Ca2+]i and is expressed as a percentage with respect to fluorescence at the beginning of the experiment. NMDA leads to an increase in [Ca2+]i that is sustained for the duration of its presence, even in experiments lasting for 10 min (not shown). In addition to the labelled cell response, a large proportion of non-labelled cells also responded to the agonist (not shown). Essentially the same results were obtained in eight out of nine labelled cells from different slices.
Figure 2 shows the effect of the non-NMDA glutamate agonist kainate in a tangentially migrating DiI labelled cell in the IZ. Figure 2A shows a low-power magnification of the slice. Figure 2B shows the DiI fluorescence image of a tangentially migrating cell from the area indicated in Figure 2A. Figure 2C shows a composite image with the Fluo-3 fluorescence (green) of the same z-stack projection captured in the previous panel. The record displayed in Figure 2D shows the time-course of the calcium response to the kainate bath applied during the time indicated by the bar. As was the case for NMDA, numerous non-labelled cells also responded to kainate. Essentially the same results were obtained in 10 other labelled cells from different slices.
Figure 3 shows the effect of the GABAa receptor agonist muscimol on [Ca2+]i in a DiI labelled cell in the IZ. The format of the figure is the same as Figures 1 and 2. Figure 3B shows the DiI fluorescence image of a tangentially migrating cell from the area indicated in Figure 3A. Figure 3C shows the colocalized fluorescence of the calcium indicator Fluo-3 (green) and DiI (red). The record displayed in Figure 3D shows the time-course of the cell calcium response when muscimol was applied to the bathing medium (indicated by the bar). Numerous non-labelled cells also responded to muscimol. In most of the experiments (see below) the response to muscimol was transitory. Essentially the same results were obtained in 15 other labelled cells from different slices.
The above results suggest that tangentially migrating cells in the IZ respond to these agonists, but do not indicate if a given cell possesses all or only some of the receptors. We therefore tested to see if the response of tangentially migrating cells was to a given neurotransmitter or to the three tested. Figure 4 shows an example of these experiments. Figure 4A shows a confocal projection of the Fluo-3 fluorescence obtained from 14 confocal images (total z = 25 μm). This preparation was especially favorable for the visualization of the MZ, CP and IZ. The arrows point to the soma and different parts of the leading process of the DiI labelled cell presented in the next panel. Figure 4B shows a DiI labelled cell in the IZ. Note the large process that reaches the MZ. The arrows are in the same position as those shown in Figure 4A. The traces displayed in Figure 4C show the changes in Fluo-3 fluorescence measured in the soma of the DiI labelled cell in response to bath application of the agonists. The first record shows the time-course of the calcium response to NMDA, which was applied to the bathing medium during the period indicated by the bar. The next trace shows the effect of kainate on the same cell and the third shows the effect of muscimol on [Ca2+]i levels in the DiI labelled cell. The gap between the three records corresponds to a period of 10–15 min, necessary for the cell to recover its initial calcium levels. Time calibration and fluorescence change calibration applies to the three records. This experiment is representative of results obtained in a total of seven cells from different slices, all of which responded to the three agonists with an increase in [Ca2+]i, with the exception of one cell that only responded to kainate and muscimol and not to NMDA.
Radially migrating cells posses functional glutamate and GABA receptors whose activation leads to [Ca2+]i changes. In order to validate our results, in a series of experiments (not shown) we placed DiI in the VZ to label radially migrating cells. These cells displayed essentially identical responses to those described above for tangential cells.
Taken together, these results indicate that tangentially migrating cells in the IZ posses functional NMDA, non-NMDA and GABAa receptors whose activation leads to changes in [Ca2+]i. We next wanted to characterize the specificity and mechanisms of the observed responses, which required quantification of the responses using ratiometric Fura-2 measurements.
Calcium Response of IZ Cells
Fura-2 experiments were initially carried out in DiI labelled cells. However, given the homogeneity of the responses, the experiments were repeated in non-labelled IZ cells in order to quantify a large number of cells. Figure 5 shows the effects of glutamate and GABA receptor activation on [Ca2+]i in IZ cells. Figure 5A shows a transilluminated image of the slice, the rectangular box indicating the area displayed in Figure 5B. Figure 5B shows a photomicrograph of the Fura-2 fluorescence excited at 340 nm, where the soma of individual cells (labelled 1–8) in the IZ can be distinguished. The traces show individual responses of these cells to consecutive bath applications of the agonists. The responses are presented as absolute changes in [Ca2+]i with respect to the resting values. Resting [Ca2+]i was measured in 383 cells from 41 different slices and had a mean value of 150 ± 2 nM (mean ± SEM). The first set of traces shows the time-course of the calcium response to NMDA. This experiment was performed in a Mg2+-free medium (see below). In 92 cells (11 slices), NMDA produced a net increase in [Ca2+]i of 160 ± 9 nM (mean ± SEM). The next set of traces show the response of the same cells to kainate. In 82 cells (10 slices), kainate induced an increase of 153 ± 9 nM (mean ± SEM). Essentially the same results were obtained with the agonist AMPA (not shown), in 64 cells (seven slices). AMPA increased the [Ca2+]i by 142 ± 10 nM (mean ± SEM). The third set of traces shows the response to muscimol. In 145 cells (14 slices), muscimol produced an increase of 202 ± 10 nM (mean ± SEM) with respect to resting levels. In other experiments (not shown), we found that neither the treatment with the metabotropic glutamate agonist DHPG, nor the GABAb receptor agonist baclofen had any effect on [Ca2+]i.
We next characterized the specificity of these responses. Figure 6 shows a summary of these results, in which the traces show averaged responses (see legend). Figure 6A shows the effects of Mg2+ on the amplitude of response to NMDA. The first trace shows the blockade of the NMDA response in a standard medium containing 1.3 mM Mg2+. The second trace shows the NMDA response from the same cells in a Mg2+-free medium. The absence of Mg2+ as a requirement for the NMDA effect (Nowak et al., 1984) reflects the characteristics of a fully mature receptor. Figure 6B shows the blocking effect of the NMDA receptor antagonist APV. The first two traces show the control response to NMDA and to NMDA in the presence of APV, respectively. The blocking effect of APV was fully reversible and a third challenge with NMDA led to a response similar to the first (not shown). Figure 6C shows the effect of the non-NMDA antagonist CNQX on the response to kainate. The first two traces show the control response to kainate and to kainate in the presence of CNQX, respectively. Figure 6D shows the effect of the GABAa antagonist bicuculline. The first two traces show the control response to muscimol and to muscimol in the presence of bicuculline, respectively. As in the case of the blockade by APV, the effects of CNQX and bicuculline were fully reversible. These results indicate that the [Ca2+]i responses observed were due to the specific activation of NMDA, non-NMDA and GABAa receptors.
Given the complexity of the slices, different cells may not be simultaneously exposed to the agonist, precluding a precise quantification of the [Ca2+]i response time-course. However, in general terms it was found that the responses to NMDA and kainate had a slow onset, taking 1–2 min to reach a plateau where the [Ca2+]i remained until the agonist was removed. In some experiments, the effect of kainate was clearly transitory (see Fig. 4). The time-course for the [Ca2+]i changes elicited by muscimol was faster, with a time-to-peak of ~30 s. The effect of muscimol was transitory and, when present for long periods (10 min), [Ca2+]i remained slightly elevated with respect to the resting level, following the initial peak (not shown).
We next investigated the ionic mechanisms implicated in the calcium increase produced by the different agonists. To examine the direct contribution of channels coupled to receptors or the depolarization-induced opening of calcium channels, we used the sodium channel blocker TTX. Figure 7 shows the effect of TTX on the agonist responses. Figure 7A shows the effect of TTX on the NMDA response. The first two traces show the control response to NMDA and to NMDA in the presence of TTX, respectively. In 32 cells, the response in the presence of TTX was 89% with respect to the control. These results indicate that the receptor activated by NMDA is coupled to a channel permeable to calcium (MacDermott et al., 1986; Schneggenburger et al., 1993). This is consistent with the lack of effect of TTX on the [Ca2+]i increase, indicating that the generation of action potentials is not necessary for the response.
It has been recently shown that cells in the IZ, with morphology of tangential migration, express Ca2+-permeable AMPA receptors (Metin et al., 2000). The high calcium permeability (Burnashev et al., 1992) correlates with the finding that the GluR2 subunit is not expressed in the IZ. In the study by Metin et al., kainate application led to an inward current that was not modified by the presence of voltage-gated sodium and calcium channel blockers. Consistent with this, Figure 7B shows that TTX has no effect on the response to kainate. The first two traces show the control response to kainate and to kainate in the presence of TTX, respectively. In 24 cells, the response in the presence of TTX was 91% of the control.
Immature cells possess a Cl− gradient that leads to an inward current (Rivera et al., 1999), the depolarization of the cell and the activation of voltage-dependent Na and/or Ca currents. Electrophysiological studies in tangentially migrating cells (Metin et al., 2000) have shown that application of GABA produces a rapidly desensitizing inward current with a reversal potential close to the Cl− equilibrium potential (–36 mV). In the same study it was found that IZ cells display inward currents that are TTX sensitive. Figure 7C shows the effect of TTX on the response to muscimol. The first two traces show the control response to muscimol and to muscimol in the presence of TTX, respectively. In 36 cells, the response in the presence of TTX was 15% of that in the control, indicating that the rise in [Ca2+]i induced by muscimol requires the generation of action potentials.
The results presented in this study show that tangentially migrating cells in the IZ of the cortex possess functional NMDA, non-NMDA and GABAa receptors whose activation leads to a rise in [Ca2+]i. These effects are due to the activation of specific receptors that seem to be ubiquitously expressed by tangentially migrating cells.
To label tangentially migrating cells, we placed DiI crystals in the MGE or LGE, close to the corticostriatal sulcus. Recent evidence (Anderson et al., 2001) indicates that cells originating in the MGE are the first to migrate through the IZ and at later stages, cells originating in the LGE migrate via SVZ and some invade the CP, resulting in a period in which both migrations are mixed. Although labelled cells in different zones of the cortex were found, we have concentrated our [Ca2+]i measurements in cells of the IZ with tangentially migrating morphological characteristics and whose origin may be the MGE.
Functional Receptors in Tangentially Migrating Cells
Tangentially migrating cells in the IZ respond to NMDA, kainate and muscimol with an increase in [Ca2+]i that can be blocked by specific antagonists. In the case of NMDA and AMPA/kainate receptor activation, the [Ca2+]i increase is due to the direct permeation of Ca2+ ions through ionotrophic receptors. In contrast, the increase in [Ca2+]i induced by GABAa receptor activation relies on membrane depolarization and the activation of voltage-gated calcium channels. There is ample evidence in the literature indicating that the activation of GABA receptors in immature neurons produces an excitatory effect (Ben-Ari et al., 1997; Ben-Ari, 2001) due to the reversal of the Cl− equilibrium potential (Rivera et al., 1999). The effects produced by the agonists tested were blocked by Zn2+ (not shown), indicating that they are due to Ca2+ entry from the extracellular space. On the other hand, agonists of metabotropic glutamate and GABA receptors had no effect on [Ca2+]i. Although our results would suggest that the effects observed are due to the direct effects of neurotransmitters in tangentially migrating cells, the possibility of indirect effects, mediated by the release of other molecules from more mature cells due to the activation of GABA and glutamate receptors should not be discounted.
In general, our results are consistent with the electrophysiological responses recorded in IZ cells with tangential morphology by Metin et al. (Metin et al., 2000), except that they did not find a response to NMDA. A possible explanation for this difference could be the age of the slices studied. In our experiments, E16 + 24–48 h slices were used, whereas the experiments of Metin et al. used E12–14 mouse embryos. It has been reported that in radially migrating cells, neurons in the VZ do not express functional NMDA receptors, whereas those in the CP do, suggesting a temporal gradient of maturation (LoTurco et al., 1991). It is conceivable that tangentially migrating cells also display a similar pattern of maturation.
Homogeneous Response of IZ Cells
One of the most striking findings of this study is the relative homogeneity of response, both within labelled tangential migrating cells and in different cells present in the IZ. The IZ is mostly composed of radially and tangentially migrating cells. In a series of experiments not included in this study, we labelled radially migrating cells and found similar responses to those displayed by tangentially migrating cells, thereby suggesting that the calcium signalling due to activation of these receptors is similar in both cell types. This may indicate possible differences in later events triggered by the rise in [Ca2+]i. For instance, recent reports (Simonian and Herbison, 2001) demonstrate the importance of NMDA receptors in the tangential migration of gonadotropin-releasing hormone (GnRH) cells, indicating that tonic NMDA receptor activation slows the final phase of migration within the forebrain. When compared with data on radially migrating cells where NMDA stimulates migration (Behar et al., 1999; Hirai et al., 1999), it would appear that NMDA activity may increase or decrease the rate of neuronal migration, depending on the system studied. Currently, there are no data about the effect of NMDA receptor activation or blockade on tangentially migrating interneurons. Likewise, despite the abundance of data on the effects of GABA on radial migration, to our knowledge, there is no information on the effect GABA on tangential migration.
There is only one report (Poluch et al., 2001) showing that the treatment of rat organotypic cultures with AMPA leads to a significant increase in the number of rounded (due to neurite retraction) MAP2 positive cells, the effect being blocked by specific antagonists. These results suggest that the activity of such receptors may participate in guiding or providing stop signals for tangential migration.
Signals for Tangential Migration
The molecular signals that guide tangential migration are largely unknown. In explant studies, it has been found that the secreted protein Slit acts as a chemorepellant (Hu, 1999; Wu et al., 1999; Zhu et al., 1999) that drives GABAergic neurons to migrate from the ganglionic eminence to the neocortex. The role of Netrin-1 has been demonstrated in the circumferential migration of basilar pontine neurons (Bloch-Gallego et al., 1999; Yee et al., 1999) and in other tangential migrations (Alcantara et al., 2000). Recently, it has been shown (Marin et al., 2001) that the sorting of striatal or cortical neurons is regulated by the semaphorin– neuropilin interaction. Thus, migrating interneurons expressing neuropilins (receptors for semaphorins) are directed toward the cortex, whereas those lacking neuropilins are directed towards the striatum. Striatal cells express Sema3A and Sema3F, creating an exclusion zone for interneurons migrating to the cortex. Sema3A has been shown to be chemorepellant for cortical axons; however, it acts as a chemoattractant for cortical apical dendrites, depending on the asymmetric localization of soluble guanylate cyclase (Polleux et al., 2000).
Ming and coworkers have recently shown that growth cone turning responses induced by gradients of attractive or repellent guidance cues are modulated by electrical activity in a Ca-dependent manner and include an enhancement of the repulsive or collapsing activity of Sema3A in cultured Xenopus spinal neurons (Ming et al., 2001). In this way, the endogenous activation of neurotransmitter receptors and subsequent [Ca2+]i rise may modulate the response of migrating cells to external cues. There is a diffuse staining for GABA and glutamate in different zones of the cortical wall, including the IZ (Haydar et al., 2000). Furthermore, it has been shown that there are growing cortical axons apposed to tangentially migrating IZ cells (Metin and Godemet, 1996) and that neurites in the IZ express a glutamate transporter (Furuta et al., 1997), suggesting that corticofugal axons may release glutamate (Metin et al., 2000). It is therefore possible that GABA and/or glutamate present in the extracellular space leads to activation of specific receptors and, through the calcium signals shown in this work, modulates tangential migration.
In conclusion, our experiments indicate that tangentially migrating cells in the IZ possess functional NMDA, non-NMDA and GABAa receptors whose activation leads to changes in [Ca2+]i. These findings suggest that at a macroscopic level these cells are quite mature and relatively homogeneous. The endogenous activation of these receptors may have either a direct effect on tangential migration or modulate the response of migrating cells to external cues.
Supported by grant SAF2000-0152-C02-02 from Ministerio de Educación y Cultura. We wish to thank Drs Pere Berbel, Costantino Sotelo and Félix Viana for their constructive criticisms on previous versions of the manuscript and Esther Ballesta for excellent technical assistance.