INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The Cajal-Retzius (C-R) cells are among the earliest generated preplate neurons of the cerebral cortex. A significant revelation has been that Cajal-Retzius cells produce and release reelin, a glycoprotein implicated in the inside-out migration of cortical neurons and the proper formation of cortical layers (Curran and D'Arcangelo 1998; D'Arcangelo et al. 1995, 1997; Ogawa et al. 1995). Thus C-R cells are in a favorable position to play an important morphogenic role in the development of the neocortex. However, information on other functional aspects of C-R cells is relatively limited. Clearly, the emergence of excitatory and inhibitory activity is critical in all phases of cortical development. Sensitivity to glutamate has been reported in rat C-R cells and this has been attributed principally to the activation of N-methyl-D-aspartate (NMDA) receptors.

In light of the role played by NMDA receptors in programmed cell death, it has been postulated that activation of NMDA receptors contributes to the disappearance of most C-R cells (Mienville and Pesold 1999), which occurs within the first two postnatal weeks of cortical development in rodent (Alcantara et al. 1998; Del Rio et al. 1995; Derer and Derer 1990; Meyer et al. 1998; Mienville and Pesold 1999). Unlike the situation in rodent, a significant population of C-R cells in primates, including that in humans, survives and persists into adulthood (Belichenko et al. 1995; Marin-Padilla 1998; Meyer et al. 1998; Zecevic and Rakic 2001). One might postulate that primate C-R cells express glutamate receptors with functional properties that differ from those in rodents. Differential expression in ionotropic glutamate receptors could account in part for the species-dependent difference in the survival of C-R cells. However, a prerequisite would be a demonstration of the presence of glutamate receptors in human C-R cells. In this study, we report for the first time glutamate-activated current responses in human C-R cells. We demonstrate, in addition to the presence of NMDA receptors, sensitivity to -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) in human but not in mouse C-R cells.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Preparation of cortical slices

Neonatal mouse pups (Swiss Webster) between 4 and 6 postnatal days old (postnatal day 0 is the day of birth) were used to prepare cortical slices. Midgestational human fetal cortical tissue (16-24 wk of gestation) was obtained from the Albert Einstein University Brain Bank. All procedures were performed in accordance with National Institutes of Health guidelines.

Thin brain slices (150-250 µm thick) containing neocortex were prepared using a vibroslice (model VSLM1, WPI, Sarasota, FL). The slices were cut in ice-cold cutting solution containing (in mM) 110 sucrose, 3 KCl, 7 MgCl₂, 1.25 NaHPO₄, 0.5 CaCl₂, 28 NaHCO₃, 5 dextrose, 0.6 ascorbate, and 0.1 kynurinate. Prior to electrophysiological recording, the slices were stored at room temperature for at least 30 min in a reservoir of oxygenated artificial cerebral spinal fluid (aCSF) consisting of (in mM) 125 NaCl, 2.5 KCl, 1 MgCl₂, 1.25 NaHPO₄, 2 CaCl₂, 25 NaHCO₃, and 25 dextrose.

Patch clamp recording and drug application

Slices were transferred to the recording chamber, where whole-cell patch clamp recording was performed with an Axopatch 200A patch clamp amplifier (Axon Instruments, Burlingame, CA). The recording chamber was tightly secured to the stage of an upright microscope and perfused continuously with oxygenated aCSF at a rate of 0.5-0.8 ml/min. All recordings were performed at room temperature.

Individual cells in cortical slices were resolved under Hoffman modulation optics using a 40× water immersion objective (3 mm working distance) (Olympus, Lake Success, NY). Real-time images were captured on-line with a charge-coupled device camera (MTI 300RC, Dage-MTI, Michigan City, IN) and displayed on a video monitor (MTI-HR 1000, Dage-MTI). This guided the navigation and placement of the recording and drug pipettes. The pipette solution used for whole-cell recording contained (in mM) 20 KCl, 120 K-gluconate, 0.5 EGTA, 10 HEPES, 4 NaCl, 2.5 MgATP, 0.25 NaGTP, and 5 QX 222. Data were acquired and analyzed using the pClamp software family (version 6.0) via a TL-1 Labmaster interface board (Axon Instruments). Lucifer yellow (0.1%) was included in the pipette solution to reveal the morphology of the recorded cells during electrophysiological recording and then to aid in identification of the recorded cells following subsequent fixation of the slice for immunohistochemical processing.

Glutamate and its agonists and antagonists were loaded into the separate barrels of a multi-barrel pipette assembly and delivered to the immediate vicinity of cells via regulated pressure controlled by a 4-channel picospritzer (General Valve Co., Fairfield, NJ). Pressure puffs of 0.5-0.8 s duration at 15 s intervals were used to apply pulses of glutamate. One of the barrels routinely contained aCSF, which was applied continuously for 14 s between glutamate applications to facilitate clearance of the agonist and to control for mechanical artifacts. An identical protocol was used to test the action of antagonists. 6-Cyano-7-nitroquinoxaline-2,3-dione (CNQX; 5 µM) (Tocris, Bellwin, MO), 2-Amino-5-phosphonovaleric acid (APV; 100 µM) (Sigma, St. Louis, MO), or ifenprodil (3 or 10 µM) (Tocris) were prepared from frozen concentrated stock and dissolved in aCSF. Current-voltage relationship of glutamate-induced responses was determined in triplicate by analyzing the amplitude and polarity of the responses at each individual membrane potential held continuously between 90 and +40 mV.

Immunohistochemistry

After each recording session, the electrophysiologically recorded slices were fixed in 4% paraformaldehyde dissolved in 0.1 M phosphate-buffered saline (PBS) for 4 h and washed subsequently three times for 5 min each in PBS. A blocking agent consisting of 1% bovine serum albumin, 5% normal goat serum, and 0.5% Triton X-100 in PBS was applied for 2 h before the slices were incubated overnight at room temperature with reelin antibody CR-50 (1:500). Following incubation with primary antibody, the slices were washed with PBS and a rhodamine-conjugated secondary antibody (1:200 dilution) was applied for 2 h. Cells displaying Lucifer yellow- and rhodamine-induced fluorescence were imaged and analyzed using a confocal microscope (LSM410, Carl Zeiss).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Human and mouse C-R cells were recorded in thin slices under patch clamp conditions to monitor whole-cell current responses activated by exogenously applied glutamate or AMPA. Figure 1 illustrates a slice from midgestational human cortex. C-R cells could be readily identified as cellular profiles in the marginal zone displaying large oval somata with long processes that often extended parallel to the pial surface (Fig. 1A, arrow). The same cell was highlighted by Lucifer yellow-induced fluorescence following patch clamp recording (Fig. 1B). Subsequent to fixation and immunohistochemical processing of the slice, this cell was re-examined and shown to express reelin immunoreactivity (Fig. 1C).

View larger version (94K): [in this window] [in a new window]   Fig. 1. A: Cajal-Retzius (C-R) cell (arrow) in a slice from midgestation human cortex, before recording. B: same C-R cell after recording and filled with Lucifer yellow (arrow). C: after fixation, the same C-R cell is shown to be immunopositive after processing with antibody against reelin (arrow). The scale bar in B also applies to C.

Current responses of human C-R cells to glutamate agonists

Figure 2A illustrates the averaged current-voltage relationship in human C-R cells (n = 3) based on current responses evoked by 100 µM glutamate. A superimposed set of traces represents current responses to glutamate at different holding potentials (Fig. 2A, inset). Glutamate induced inward current responses that decreased in amplitude at holding potentials more hyperpolarized than 30 mV. This voltage dependence resembled that typically reported for activation of NMDA receptors (Mayer et al. 1984; Nowak et al. 1984). However, we found that the NMDA receptor antagonist APV (100 µM) had minimal effect (<10%) on blocking the glutamate-induced current responses (data not shown). In recombinant NMDA receptors, sensitivity to APV appears to be subunit-dependent, insofar as NR2B-containing recombinants are less sensitive to antagonism by APV than are their NR2A-containing counterparts (Kutsuwada et al. 1992; Meguro et al. 1992). We therefore used ifenprodil, an NR2B subunit-selective antagonist, to assess the participation of the NR2B subunit in the glutamate response. Ifenprodil, tested at a concentration (3 µM) that is specific for blocking NR2B-containing recombinant NMDA receptors (Williams 1993), reduced the peak amplitude of the glutamate response by 33.4 ± 1.9% (mean ± SD; n = 3) (Fig. 2B). In contrast to what was observed in mouse C-R cells (see RESULTS below), application of AMPA (100 µM) evoked small but clear-cut current responses in six of six human C-R cells (Fig. 2C). Thus the whole-cell glutamate-induced current response in human C-R cells has both NMDA and AMPA receptor-mediated components.

View larger version (26K): [in this window] [in a new window]   Fig. 2. Glutamate-induced current responses in human C-R cells. A: voltage dependence of glutamate response. Inset: superimposed traces illustrating glutamate-induced current response at various holding potentials. The peak amplitude of the current response monitored at each holding potential was expressed as a percentage of that monitored at a holding potential of +40 mV. B: ifenprodil-induced suppression of glutamate response. C: example of an AMPA-evoked current response monitored at 60 mV holding potential in a human C-R cell.

Current responses of mouse C-R cells to glutamate agonists

As was demonstrated in human C-R cells, the glutamate-induced current responses of mouse C-R cells showed voltage dependence (Fig. 3A, inset). Application of APV (100 µM) reduced the amplitude of the glutamate response by 22.5 ± 3.7% (n = 11; Fig. 3B). This partial blockade by APV suggests that the whole-cell current response to glutamate includes an NMDA receptor-mediated component. Consistent with this notion, ifenprodil tested at 3 µM and 10 µM reduced peak amplitude of the glutamate-induced current response by 15.0 ± 3.7% (n = 7) and 30.9 ± 2.9% (n = 6), respectively (Fig. 3C). However, as illustrated in Fig. 3D, the glutamate response of mouse C-R cells was unaffected by exposure to the AMPA receptor antagonist CNQX (5 µM) (8 of 8 cases; Fig. 3D). This suggests that AMPA receptors do not participate in shaping the whole-cell glutamate response. Indeed, mouse C-R cells (8 of 8) were characteristically insensitive to AMPA (Fig. 3D) even when the agonist was applied at a high concentration (200 µM). In the same brain slices, AMPA (100 µM) could be shown to elicit robust current responses in layer II/III pyramidal cells (Fig. 3E). These observations are in distinct contrast to our findings on human C-R cells. Taken together, they suggest that AMPA receptors are absent in mouse C-R cells and that NMDA receptors are the predominant ionotropic glutamate receptor expressed by mouse C-R cells within the window of postnatal development examined.

View larger version (23K): [in this window] [in a new window]   Fig. 3. Glutamate-induced current responses in mouse C-R cells. A: voltage dependence of glutamate response. Inset: superimposed traces illustrating glutamate-induced current response at various holding potentials. The peak amplitude of the current response monitored at each holding potential was expressed as a percentage of that monitored at a holding potential of +30 mV. B: amino-5-phosphonovaleric acid-induced partial block of the glutamate response in mouse C-R cells. C: ifenprodil (10 µM) partially blocked glutamate-induced current response. D: mouse C-R cells were insensitive to -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA). 6-Cyano-7-nitroquinoxaline-2,3-dione did not attenuate the glutamate response. E: AMPA-activated response in a layer II/III pyramidal neuron. B and D are recordings from the same C-R cell.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study examined glutamate-activated current responses in human and mouse C-R cells, focusing on the ionotropic glutamate receptors. A major finding is that AMPA, a non-NMDA receptor agonist, elicits responses from human but not mouse C-R cells. We conclude that C-R cells found in midgestational human cortex express AMPA receptors and that these receptors are demonstrably absent in C-R cells during the first postnatal week in rodent. The implication of an apparent differential expression of AMPA receptor can be speculated, at best. It may be that co-expression of AMPA and NMDA receptors in human C-R cells facilitates depolarization-induced activation of NMDA receptors at a time in development when this process is absent in mouse C-R cells. Cellular events downstream from NMDA receptor activation may contribute to regulating the activity of C-R cells, such as firing pattern or expression of reelin, in orchestrating the formation of neuronal circuits and lamination in the developing neocortex. In so speculating, a prerequisite assumption is the presence of glutamate in the extracellular milieu of the marginal zone to activate the glutamate receptors. Spontaneous postsynaptic currents have been recorded in mouse C-R cells in neonatal cortical slices (Lu and Yeh, unpublished observations). In rat, although the spontaneous postsynaptic currents monitored in C-R cells are action potential-independent and have been attributed to mediation by GABA_A rather than glutamate receptors (Kilb and Luhmann 2001), Schwartz et al. (1998) presented evidence arguing in favor of at least the large C-R cells expressing a number of synaptically functional receptor types, including glutamate receptors. Because C-R cells display immunoreactivity to glutamate (Del Rio et al. 1995), it is possible that C-R cells release the transmitter. Alternatively, newly derived cortical plate neurons may release glutamate spontaneously, establishing an ambient level of extracellular glutamate as they migrate and reach the superficial layers of the developing cortex. Thus release through synaptically mediated mechanisms and an ambient extracellular level of glutamate both are possible sources of glutamate that are not mutually exclusive.

Our conclusion does not rule out the possibility that AMPA receptors may be present in mouse C-R cells at earlier or later time points in the course of cortical development, nor that they may be expressed in a small subpopulation of C-R cells, as has been reported in rat by a number of studies (Kim et al. 1995; Mienville 1999; Mienville and Pesold 1999; Schwartz et al. 1998). Because the first postnatal week in rodent approximates the third trimester in human gestation (West and Pierce 1986), the expression of AMPA receptors may simply become manifest earlier in human than in rodent C-R cells. In this light, we consider our finding preliminary, insofar as future experiments will need to address systematically the issue of temporal expression of the ionotropic glutamate receptor subtypes in C-R cells. In the same vein, there is a need to examine the developmental expression of metabotropic glutamate receptors that have been reported to be present in C-R cells of neonatal mouse cortex (Martínez-Galán et al. 2001).

Both in human and in rodent C-R cells, the glutamate-induced whole-cell response exhibits strong voltage dependence, which suggests the involvement of NMDA receptors. The results of pharmacological tests employing ifenprodil and APV to determine receptor specificity provide additional supportive evidence. One of the issues that needs to be resolved is whether the developmental expression profile of NMDA receptor subunits in human and rodent C-R cells differs. For example, there is evidence that the switch from the prenatally prevalent NR2B subunit to the NR2A subunit typical of mature neurons in the cortex and hippocampus (Monyer et al. 1992, 1994; Sheng et al. 1994; Watanabe et al. 1992) may not occur in rodent C-R cells (Mienville 1999). The predominant persistence of the NR2B subunit in the makeup of NMDA receptors has been proposed as one possible factor leading to the death of C-R cells in rodents (Mienville and Pesold 1999). Given that primate and human C-R cells endure (Belichenko et al. 1995; Marin-Padilla 1998; Meyer et al. 1998; Zecevic and Rakic 2001), it is tempting to speculate that they express glutamate receptors with subunits that, on balance, do not favor cell death. With regard to the NMDA receptors, whether there is a species-dependent difference in the switch between the NR2A and NR2B subunits during cortical development, and whether such a difference may contribute to the death or survival of C-R cells, awaits further elucidation.