|Home | About | Journals | Submit | Contact Us | Français|
To investigate developmental changes in neuro-steroid modulation of GABAA receptors, whole-cell currents were elicited by applying GABA with allopregnanolone or pregnenolone sulfate (PS) to dentate granule cells (DGCs), acutely isolated from 7–14-day-old and adult rats. GABA evoked larger currents from dentate granule cells acutely isolated from adult rats (adult DGCs) than from neonatal DGCs, due to increased efficacy (1662±267 pA in adult DGCs versus 1094±198 pA in neonatal DGCs, P=0.004), and current density (0.072±0.01 pA/μm2 in neonatal rat DGCs to 0.178±0.02 pA/μm2 in adult DGCs), but unchanged potency (EC50 was 18.532 μM in adult DGCs, and 26.6±7.9 μM in neonatal DGCs, P=0.21). Allopregnanolone sensitivity of GABAA receptor currents increased during development due to an increased potency (21.1±4.7 nM in adult DGCs versus 94.6±9 nM in neonatal DGCs, P=0.0002). The potency and efficacy of PS inhibition of GABAA receptor currents were remained unchanged during development (13±6 μM and 13.2±5.9 μM, P=0.71 and 85.5%±3.5% and 83.6%±0.8%, P=0.29, respectively). To investigate possible mechanism of developmental changes in GABAA receptor properties, in situ hybridization for α1, α4 and γ2 subunit mRNAs was performed in dentate gyrus of the two age groups. Qualitatively, α1 subunit mRNA was expressed at low levels in neonatal rats while it was well expressed in adult rats. The α4 and γ2 subunits were well expressed in the dentate gyrus of adult and neonatal rats. Immunohistochemical staining for α1 subunit in hippocampal slices from neonatal and adult rats was examined under confocal laser scanning microscope. This demonstrated that cell bodies and dendrites of granule cells are moderately positive for the α1 staining in adult rats but weakly so in neonatal rats. Higher-magnification images demonstrate large number of clusters of α1-subunit in the cell bodies of dentate granule cells of adult rat but rare clusters in granule cells of neonatal rats.
Maturation of GABAA receptors in DGCs is characterized by increased number of GABAA receptors that are more sensitive to endogenous neurosteroid allopregnanolone, which might be related to increased expression of α1 subunit.
GABA, a major inhibitory neurotransmitter in the forebrain, exerts its actions by activating GABAA receptor and GABAB receptors. Fast inhibitory neurotransmission in the brain is primarily mediated by GABAA receptors, which are composed of subunits derived from the members of six gene families, α, β, γ, δ, ε, and π. Multiple different types of GABAA receptors can be assembled from these sub-units, and these subunits impart unique properties upon the receptors. There is growing evidence that the properties and subunit composition of GABAA receptors change during normal development of the nervous system.
Developmental changes in GABAA receptors can be studied effectively in dentate granule cells (DGCs). Most DGCs are born, migrate and mature in the postnatal period and neurogenesis persists during the entire lifespan of the animal (Kaplan and Hinds, 1977; Cameron et al., 1993). Levels of α1, γ2 and δ subunit mRNA, as well as corresponding polypeptide products, increase during maturation of DGCs (Wisden et al., 1992; Fritschy et al., 1994; Brooks-Kayal et al., 1998, 2001). The properties of GABA-mediated inhibitory postsynaptic currents (IPSCs) change substantially during postnatal development and are correlated to changes in subunit expression. GABAA receptor-mediated miniature IPSCs recorded from DGCs of newborn rats have slower frequency and slower rise- and decay-time constants compared with those recorded from DGCs of 21-day-old animals (Hollrigel and Soltesz, 1997). Zinc and furosemide inhibit GABAA receptor currents more strongly in neonatal DGCs than in adult DGCs. In addition, in DGCs of adult rats, zolpidem and diazepam enhancement of GABA currents is stronger than in DGCs of neonatal rats (Kapur and Macdonald, 1999).
Neurosteroids are a group of compounds synthesized in the brain from circulating steroids or de novo from cholesterol, which has anxiolytic, hypnotic, anesthetic and anticonvulsant actions. Allopregnanolone, a metabolite of progesterone, potently enhances GABAA receptor/chloride-channel function (Lambert et al., 1995; Majewska, 1992). Sulfated ester of pregnenolone, pregnenolone sulfate (PS), inhibits GABA-mediated currents in micromolar concentrations (Majewska et al., 1988; Wu et al., 1991). The neurosteroid modulation of GABAA receptors depends on the subunit composition of the receptor (Zhu et al., 1996b; Maitra and Reynolds, 1998, 1999; Shingai et al., 1991; Brussaard et al., 1997; Smith et al., 1998). Because the subunit composition of GABAA receptors on dentate granule cells changes during postnatal development, it is likely that neurosteroid sensitivity of GABAA receptors on DGCs changes during development.
All procedures on animals were performed according to the protocol approved by the Institutional Animal Care and Use Committee. The minimal number of animals necessary to conduct the experiments were used. All efforts were made to minimize distress and suffering of the animals. The animals were maintained at artificial 12-h dark/light cycle and had free access to food and water. Fourteen adult and 13 neonatal pups were used for the experiments. Adult rats were housed individually in separate cages and the pups remained in the cage with their littermates and mother. DGCs were acutely isolated from adult rats (adult DGCs) or 7–14-day-old rat pups (neonatal DGCs) according to a protocol of Kay and Wong (Kay and Wong, 1986) as modified (Kapur and Coulter, 1995). The rats were decapitated under halothane anesthesia, the brains were dissected free, chilled to 4 °C for 1 min in a piperazine-N,N′-bis(2-ethanesulphonic acid) (PIPES) buffer solution composed of (in mM): NaCl 120, KCL 5, CaCl2 1.5, MgCl2 1, D-glucose 25, PIPES 20 (pH 7.0) (all chemicals were from Sigma, St. Louis, MO, USA unless noted otherwise) and the region containing hippocampus was blocked. The brain was mounted on a vibratome stage (Camden Instruments, UK) and cut into 400–500-μm coronal sections containing hippocampus. Hippocampal sections were incubated in Sigma type XXIII protease enzyme from Aspergillus oryzae in the PIPES-buffered solution at 32 °C for 30–45 min. After a minimum of 30 min of recovery the dentate gyrus of the hippocampus was dissected from the rest of the section under a microscope and cut into 0.5-mm3 chunks. These chunks were triturated through fire-polished glass pipettes and the isolated neurons were plated on 35-mm plastic Petri dishes. Granule cells were identified by their typical oval shape and single process.
Whole-cell GABAA receptor currents were recorded from acutely isolated DGCs as described in the past (Kapur and Macdonald, 1996). The neurons were studied on the stage of an inverted microscope (Nikon Diaphot) at room temperature, with patch electrodes were filled with a recording solution containing (in mM): Trizma phosphate (dibasic) 110, Trizma base 28, EGTA 11, MgCl2 2, CaCl2 0.5: pH 7.40, the osmolarity was 270–275 mOsm. ATP disodium salt in final concentration of 3 mM was included in intracellular solution before recording. The Tris phosphate-based internal solution was used because strong buffering properties of Trizma allow inclusion of ATP into the internal solution without affecting the pH. The extracellular recording medium contained (in mM): NaCl 155, KCl 3, MgCl2 1, CaCl2 3 and HEPES-Na+10; pH 7.4, the osmolarity was 318–322 mOsm. Patch electrodes were pulled on a P-97 Flaming-Brown horizontal puller (Sutter Instruments, CA, USA) using a two-stage pull from thin-walled borosilicate glass capillary tubes (1.5-mm OD, 1.1-mm ID, World Precision Instruments, Sarasota, FL, USA) to a final resistance of 5–8 MΩ. Electrodes were fire polished as necessary. The neurons were voltage-clamped to 0 mV and thus, the application of GABA produced outward Cl−currents. The currents were amplified (Axo-patch 200A, Axon Instruments, Union City, CA, USA) and low-pass filtered at 3 kHz with an eight-pole Bessel filter prior to digitization, storage and display using patch-clamp technique (Hamill et al., 1981). The currents were displayed on-line on a Gould Viper TA11 digital chart recorder (Gould instruments, Valley View, OH, USA), and on a Pentium II personal computer using Axotape program (Axon Instruments). The currents were digitized at 400 Hz. The peak currents were measured manually from the computer and chart paper recording.
Allopregnanolone (Tocris, Ballwin, MO, USA) was dissolved in dimethyl sulfoxide (DMSO) upon arrival in the laboratory and the stock solutions were stored at −20 °C and diluted in extracellular solution in the day of an experiment. DMSO was diluted 1:50,000 or more. In control experiments, this dilution of DMSO applied to cells did not evoke any response. All the other chemicals were obtained from Sigma. GABA and other drugs were dissolved in the extracellular solution and applied to the neurons with a modified U-tube “multipuffer” application system (Greenfield and Macdonald, 1996). The concentration-response data for each group were fit to four-parameter logistic equation (equation for a sigmoid curve): I=I(max)/(1+10^((log EC50−log[GABA])×Hill slope)), where I is the peak GABAA receptor current at a given GABA concentration. The Hill slope, EC50 and I(max) (maximal current) were derived from the equation that best fit the observed (pooled) data by least square fit method using the GraphPad Prism 3.0 curve-fitting program on an IBM PC compatible computer. All values are reported as means±S.E.M.
In situ hybridization was performed according to techniques described in the past (Harrison et al., 1996). Animals were killed by deep anesthesia followed by rapid decapitation, their brains removed and frozen on dry ice. Coronal brain sections (20 μm) were cut on a cryostat and mounted onto charged slides (Fisher Scientific, Pittsburgh, PA, USA), and stored at −80 °C. On the day of hybridization, sections were fixed with for 5 min in 4% paraformaldehyde, rinsed extensively in phosphate-buffered (0.1 M) saline (PBS). In order to minimize non-specific binding, slides were treated with 0.2% glycine (in PBS) as well as 0.25% acetic anhydride (prepared fresh in 0.1-M triethanolamine-buffered saline, pH 8). The tissue was dehydrated and delipidated in graded series of ethanols and chloroform, and air-dried prior to hybridization. Oli-gonucleotide cDNA probes were custom synthesized by Operon Technologies (Almaeda, CA, USA). Probe sequences for α1, α4 and γ2 subunits were identical to those used by others in the past (Sperk et al., 1998; Wisden et al., 1991; Wisden et al., 1992). cDNA probe was [33P] labeled by enzyme terminal deoxyneucleotidyl transferase (Gibco, Invitrogen Corporation, Carlsbad, CA, USA) and [33P]dATP (New England Nuclear, Welleslay, MA, USA) at 37 °C. The labeled probe was separated by centrifugation in micro-columns. The probe was diluted to 40 million cpm/ml in a hybridization solution consisting of 50 formamide, 10% dextran sulfate, 1× Denhardt’s (0.02% each of ficoll, polyvinylpyrrolidone, and bovine serum albumin), 4× SSC (1× SSC: 150-mM sodium chloride, 15-mM sodium citrate, pH 7), 100-mM dithiothreitol, 0.5 mg/ml salmon testis DNA and 0.25 mg/ml yeast tRNA. Sections were incubated with hybridization solution, 75 μl/slide, overlaid with glass coverslips, overnight at 37 °C in a chamber humidified by 50% formamide. Following hybridization, coverslips were removed and sections were dipped twice in 1× SSC at 55 °C and then incubated at 55 °C for 60 min in 1× SSC at 55 °C. Subsequently they were incubated at 37 °C in 1× SSC for 60 min. SSC buffer for all washes contained 10-mM sodium thiosulfate. Slides were exposed to film (hyperfilm, βMax, Amersham) for 3–10 days. Sections from 7 to 14 days old and 50-day-old rats were processed simultaneously.
Male and female pups of the Sprague–Dawley strain were used. Brains were collected from six pups at postnatal day (P) 12; the day of birth of the offspring was taken as P0. Pups were deeply anesthetized with sodium pentobarbital and perfused intracardially with 0.9% NaCl, followed by 4% paraformaldehyde in 0.1-M phosphate buffer (PB, pH 7.4). The brains were removed and postfixed for 1 h in the same fixative at 4 °C. After incubation in 30% sucrose in 0.1-M PB for cryo-protection overnight, the brains were frozen by immersion in −70 °C isopentane and sectioned on a cryostat. Four coronal sections from each pup (40 μm) containing dorsal hippocampus were processed for free-floating immunohistochemistry.
Following three washes in 0.1-M PB and preincubation with blocking solution containing 10% normal goat serum (Jackson Immunoresearch Laboratories, West Grove, PA, USA) for 1 h, they were incubated with anti-GABAA receptor subunit α1 (Alomone Laboratories, Israel, 1.5 μg/ml) at 4 °C for 72 h on a shaker. The sections were then incubated in goat anti-rabbit conjugated with Alexa Fluor 488 (Molecular Probes, Eugene, OR, USA) for 60 min on a shaker at room temperature in darkness. Each step was followed by three 15-min washes with PBS. Sections were then mounted on slides with Gel/Mount (Foster City, CA, USA), the edge of each coverslip was sealed with clear nail polish and slides were stored at −20 °C. The primary and secondary antibodies were diluted in 0.1-M PBS (pH 7.4) containing 2% normal goat serum, and 1% bovine serum albumin (Jackson Immunoresearch Laboratories).
Sections were studied on TE-200 epifluorescence microscope (Nikon, Inc., Japan) equipped with an argon ion laser. Images were acquired with an Orca-1 Hamamatsu CCD camera. Confocal- and camera-based image acquisition and processing were driven by SimplePCI software (version 4.0.6, Compix, Inc., Japan). Alexa 488 fluorochromes were visualized by the argon laser (488 nm). All confocal microscopy was performed at the W. M. Keck Cell Imaging Center at the University of Virginia. Each hippocampal section was examined with low-magnification objective lens (20×) to locate dentate gyrus and granule cell layer by their morphology. High-resolution digital images were acquired using a 40× (NA 1.0) and a 100× objective (NA=1.4), and were processed in PhotoShop 6.0 (Adobe, San Jose, CA, USA). The brightness and contrast of the images were adjusted so that only punctate fluorescence, but no weak diffuse background labeling, was visible (Scotti and Reuter, 2001).
GABA (10 μM) elicited larger currents in adult DGCs (924±203 pA, n=5) than in neonatal DGCs (330±65 pA, n=7, P=0.009, t-test, Table 1). The developmental changes in GABA sensitivity of GABAA receptors were characterized in detail by obtaining a concentration-response curve. Multiple concentrations of GABA (1 μM–3 mM) were applied to neonatal DGCs and adult DGCs. In both groups of cells, peak amplitude of GABA-evoked currents increased in a concentration-dependent fashion (Fig. 1A, B). The potency (EC50) and efficacy (maximal enhancement) of GABAA receptors in both groups of cells were obtained by fitting the concentration-response data to an equation for sigmoid curve. Efficacy increased with age, in neonatal DGCs the maximal GABA-evoked current was 1094±198 pA, whereas in adult DGCs it was 1662±267 pA (P=0.004, t-test, Fig. 1C). These findings of larger GABAA-receptor currents elicited from adult DGCs are similar to those of previous studies (Hollrigel and Soltesz, 1997; Kapur and Macdonald, 1999).
Increased peak current amplitude in adult DGCs was due to increased current density. The current density in DGCs was measured by dividing the maximal current evoked by GABA (3 mM) by the membrane capacitance. The membrane capacitance was measured by the whole-cell capacitance compensation potentiometer on the am-plifier. In the neonatal DGCs, the average capacitance was 15.6±0.86 pF (n=4) and in adult DGCs it was decreased to 9.3±1.69 (n=5, P=0.007, unpaired t-test, Table 1). Assuming that capacitance/membrane-area relationship was 0.01 pF/cm2, the current density increased from 0.072±0.01 pA/μm2 in neonatal DGCs to 0.178±0.02 pA/μm2 in adult DGCs (P=0.007, unpaired t-test, Table 1). The potency of GABA in neonatal DGCs (EC50=26.6±8.9 μM, Hill slope=0.93, n=9) was not sig-nificantly different from that in the adult DGCs (EC50=18.5±2 μM, Hill slope=1.1, n=11, P=0.21, t-test, Fig. 1C, Table 1).
In adult DGCs, 10-nM allopregnanolone robustly augmented 10-μM GABA-evoked currents, 46.3±6.3% (n=4), but in neonatal DGCs the augmentation was minimal, 13.2±2.4% (n=6, P=0.0003, unpaired t-test, Table 1, Fig. 2A, B). The threshold concentration of allopregnanolone for enhancement of GABA-elicited current enhancement was lower in adult DGCs than in neonatal DGCs. Allopregnanolone (3 nM) minimally enhanced 10-μM GABA-evoked current in neonatal DGCs: 0.25±2.7 (n=7), while in adult DGCs it potentiated them by 24.9%±7.8 (n=5, P=0.006. unpaired t-test, Table 1). The threshold concentration for enhancement in adult DGCs was 300 pM.
The allopregnanolone concentration, GABAA-receptor-current augmentation relationship was studied in adult and neonatal DGCs to determine whether the stronger allopregnanolone enhancement of GABAA receptor currents in adult DGCs was due to increased potency (EC50) and/or efficacy of allopregnanolone. Multiple concentrations of allopregnanolone, ranging from 3 nM to 1 μM, were co-applied with 10-μM GABA to neonatal DGCs (n=9) and adult DGCs (n=5). Concentrations of allopregnanolone higher than 1 μM were not used because they directly activate GABAA receptors (Twyman and Macdonald, 1992). Allopregnanolone enhanced GABA-elicited currents in both groups of DGCs in a dose-dependent fashion. The data from the two groups of cells were fit to an equation for sigmoidal curve (see Experimental Procedures) to derive potency. In neonatal DGCs EC50 of allopregnanolone was 94.7±9 nM, Hill slope=0.86 (n=9) and in adult-rat granule cells it was 21.16±4.7 nM, Hill slope+1.22 (n=5, P=0.0002, t-test, Fig. 2C, Table 1). Thus, allopregnanolone was less potent in enhancing of GABAA receptor currents in neonatal DGCs than in adult DGCs. A plateau of allopregnanolone-concentration response curve was not reached in neonatal DGCs due to direct effects of allopregnanolone, and therefore, the efficacy of allopregnanolone could not be compared.
Thirty-micromolar PS robustly inhibited 30-μM GABA-evoked currents in adult DGCs by 69.5%±0.5 (n=3) and in neonatal DGCs by 61.4%±7.3 (n=7, P=0.58, unpaired t-test). PS-mediated inhibition of GABA currents was further studied by obtaining detailed concentration-response curves from the cells isolated from both age groups. PS was co-applied with 30-μM GABA in concentrations ranging from 300 nM to 300 μM. PS inhibited GABAA receptor-mediated currents in concentration-dependent manner in dentate granule cells isolated from hippocampi from both age groups (Fig. 3A, B). In adult DGCs the IC50 was 13±6 μM (n=8) and in 7–14-day-old rat DGCs IC50 was 13±6 μM (n=5, P=0.71, Table 1, Fig. 3C). Maximal inhibition of GABA-evoked currents by PS in 7–14-day-old DGCs was 83.6%±0.8 (n=3) and 85.5%±3.5 (n=6) in adult DGCs (P=0.29, t-test, Fig. 6, Table 1). Thus, the potency and efficacy of PS inhibition of GABAA receptor-mediated currents in hippocampal dentate granule cells does not undergo significant changes between these two age groups.
Previous studies on GABAA receptors demonstrated that allosteric modulation of GABA function by allopregnanolone is markedly increased in the presence of α1 subunit (Shingai et al., 1991; Brussaard et al., 1997) and decreased when α4 subunit expression is increased (Smith et al., 1998). We detected α1, α4 and γ2 subunit mRNA expression in the hippocampus of 7–14-day-old rats and adults using in situ hybridization. In the dentate gyrus and pyramidal cell layer of the hippocampus of neo-natal rats (n= 4) (Fig. 4A, top panel), the α4 and γ2 subunit mRNA were well expressed but α1 subunit mRNA was poorly expressed. In adult rats (Fig. 4A, lower panel), these probes revealed that all three transcripts were well expressed in the dentate gyrus and pyramidal cell layer of the hippocampus of adult rats, a pattern similar to that reported by other laboratories (Sperk et al., 1998; Laurie et al., 1992). These findings were replicated in three separate experiments. These results were obtained from three sections, each from three neonatal and three adult rats.
Radioisotopic in situ hybridization has low spatial resolution, so it is not possible to localize mRNA expression to specific cell types or to cell body or dendrites. Interneurons on the hilar border of dentate gyrus express α1 subunit, and mRNA from these cells contribute to the in situ hybridization signal. In addition the correspondence between GABAA-receptor subunit mRNA and polypeptide expression may not be one to one. In order to overcome these limitations coronal sections from the hippocampus of neonatal and adult rats were labeled with antibodies against α1 subunit of the GABAA receptor. The detected primary antibody was labeled with fluorescent-labeled secondary antibody and the resulting fluorescence was using CLSM (Fig. 5). The dentate gyrus was identified using 20× objective and images of dentate gyrus were acquired with 40× (Fig. 5A, B). The distribution of α1-subunit staining in cell bodies of granule cells was investigated in detail in images acquired with 100× objective (Fig. 5C, D).
In low-magnification images obtained with 40× objective (Fig. 5B), the distribution of α1-subunit polypeptide in dentate gyrus of adult rats was similar to that described in the past (Esclapez et al., 1996) In the granule cell layer, the cell bodies of granule cells were moderately labeled for α1 subunit. Punctate staining was present along the dendrites of granule cells in the molecular layer. Interneurons on the hilar border were intensely labeled for the α1 subunit. In contrast, in sections from neonatal rats staining for the α1 subunit was weak or absent in dentate granule cells bodies (Fig. 5A). The dendrites in the molecular layer were moderately stained for α1 immunoreactivity. In sections from neonatal rats interneurons on the hilar border were as intensely stained in neonatal hippocampus as in adult hippocampus. Low level of immunostaining of dentate granule cells in neonatal rats is unlikely to be due to failure of the immunohistochemical procedure. This finding was consistently present in all sections from three pups. In addition, same sections that showed low levels of α1 immunoreactivity in granule cells showed more intense staining of interneurons.
The paucity of α1 immunoreactivity in the cell bodies of granule cells of neonatal rats was particularly striking, and was examined in greater detail under higher magnification with a 100× objective. In hippocampal sections from adult rats clusters of α1 immunoreactivity outlined cell bodies of dentate granule cells (Fig. 5D). In contrast, cell bodies of dentate granule cells in sections from neonatal rats were largely devoid of α1-immunoreactive clusters (Fig. 6C). Low-level diffuse staining for α1 subunit was present in cell bodies of some granule cells and occasional puncta were also noted. The interneurons were intensely positive for α1 in sections from neonatal and adult rats (lower right corner Fig. 5C and D).
There were three findings in this study. First, the efficacy of GABA increases during postnatal development of DGCs. Second, the allopregnanolone enhancement of GABA-evoked currents increased but the PS inhibition of GABA-evoked currents did not change during the postnatal development of DGCs. Finally, the expression of α1-subunit mRNA in the dentate gyrus and α1-subunit polypeptide immunoreactivity in the dentate granule cells increased during postnatal development.
The allopregnanolone sensitivity of GABAA receptors on dentate granule cells increased during postnatal development. Several binding studies published in the past support these physiological data. Allopregnanolone enhancement of binding of [3H]GABA to synaptosomes increases during development in chick optic lobe (Viapiano et al., 1998) and rat brain cortex (Borodinsky et al., 1997). Allopregnanolone potentiation of [3H]muscimol and [3H]flunitrazepam binding to membranes isolated from guinea-pig neocortex was found to be increasing during postnatal development (Bailey et al., 1999).
In contrast to the current study, a previous study reported a developmental decrease in neurosteroid sensitivity (Cooper et al., 1999) of GABAA receptors on DGCs. Tetrahydrodeoxycorticosterone (THDOC) prolonged slow decay-time constants of miniature IPSCs recorded from DGCs in 10-day-old rats, but not in 20-day-old animals. Several methodological differences may account for the differing results of the two studies. Neurosteroid modulation of GABAA receptors was studied here using whole-cell currents evoked by application of GABA to isolated neurons. These whole-cell currents are comprised of currents mediated by both synaptic and extrasynaptic GABAA receptors. In contrast, Cooper et al. (1999) studied neurosteroid modulation of GABAA receptor miniature IPSCs, which are mediated by the synaptic GABAA receptors. A growing number of studies suggest that kinetic and pharmacological properties of GABAA receptors clustered at synapses are different from those located on extrasynaptic membrane. Thus, in cerebellar granule cells, δ subunit was predominantly found on extrasynaptic dendritic and somatic membrane (Nusser et al., 1998). Predominantly extrasynaptic δ subunit-containing GABAA receptors do not desensitize after prolonged application of GABA and α6 and δ subunit-containing receptors have high affinity to GABA (Saxena and Macdonald, 1996). Therefore, participation of extrasynaptic receptors in whole-cell GABA-evoked currents can yield a picture of allopregnanolone sensitivity, different from that where only synaptic receptors are involved.
In addition, increased sensitivity to allopregnanolone was demonstrated at multiple concentrations of the steroid, while in the study of Cooper et al. (1999), the effect was observed at single 100-nM concentration of THDOC. Finally, in the current study, allopregnanolone potentiation of GABAA receptor currents was characterized using low, 10-μM concentration of GABA. In this experimental paradigm, potentiation by very low concentration of allopregnanolone can be detected reliably. In contrast, in the study of Cooper et al. (1999) THDOC effects were studied on synaptic currents, which are produced by transiently released millimolar concentrations of GABA (Clements, 1996; Jones and Westbrook, 1995, 1996), and thus changes in allopregnanolone sensitivity described in the current paper may not be detected in the presence of saturating levels of GABA.
In addition to allopregnanolone sensitivity of GABAA receptors, the allopregnanolone synthesis is also developmentally regulated. Expression of 5α-reductase type II mRNA, an isoform of allopregnanolone-synthesizing enzyme in the brain, is developmentally regulated (Poletti et al., 1998a. RT/PCR analysis of embryonic and neonatal rat brain mRNAs demonstrated that type II mRNA was transiently expressed only at the end of gestation (embryonic day 18) through the early postnatal period, dramatically falling by P14. This pattern of expression of type II mRNA correlated with testosterone synthesis in the fetal testis and may also suggest that 5α-reductase type II may be involved in regulating brain sex differentiation at a critical period. Transient expression of 5α-reductase mRNA was found in the proliferative zones of the developing CNS, suggesting that 5α-reduced steroids may have a role in neuronal proliferation (Compagnone and Mellon, 2000). Dehydroepiandrosterone and dehydroepiandrosterone sulfate were shown to regulate the motility and/or growth of neocortical neurons in primary cultures of mouse embryonic neocortical neurons, at concentrations normally found in the brain (Compagnone and Mellon, 1998).
The current study demonstrates that GABA efficacy increases during postnatal development of DGCs. Similar findings have been reported in the past. During postnatal development of DGCs, the frequency and amplitude of miniature postsynaptic currents (mIPSCs) increase (Hollrigel and Soltesz, 1997). The similar pattern of increased frequency of GABA-ergic mIPSCs and decrease of decay time constant was observed in cortical neurons (Dunning et al., 1999). A developmental increase in a whole-cell GABAA-receptor currents recorded from DGCs was also reported in the past (Kapur and Macdonald, 1999). This increase in GABA efficacy during development is likely due to a larger number of GABAA receptors expressed by the adult DGCs (Kapur and Macdonald, 1999) since the current density also increased during postnatal development. Increased current density observed in this study was similar to that reported in acutely isolated thalamic and cortical neurons (Oh et al., 1995; Gibbs et al., 1996).
GABAA receptor antagonist properties of PS had been described before (Majewska et al., 1988; Lambert et al., 1996; Shen et al., 1999). The present study demonstrates that micromolar concentrations of PS are equally effective in inhibition of GABAA-receptor currents in neonatal and adult DGCs. Different mechanisms and sites of action of allopregnanolone and PS on the GABAA receptors may explain this distinction between actions of allopregnanolone and PS (Park-Chung et al., 1999). In a study of the relationship between structure of various neuroactive steroids and their effect on GABA-evoked currents, it was found that the positive and negative modulators act at different sites, and can independently regulate GABAA receptors. Studies on recombinant GABAA receptors with known subunit composition demonstrate that allopregnanolone and PS do not compete for the same site (Park-Chung et al., 1999). Several ligand-binding studies have also reached similar conclusions (Gee et al., 1988, 1989).
A proliferative zone in adult dentate gyrus generates DGCs throughout the life. Little is known about the role of these newly generated DGCs. Our results raise a question whether the diminished efficacy of GABA and the potency of allopregnanolone in neonatal DGCs is a function of age of an animal or is a property of newly generated DGCs. Although this subject exceeds the scope of the present paper and would require a special study, some preliminary results obtained by others indicate that newly generated DGCs, which make up small percentage of the total population of DGCs, undergo normal differentiation in the adult brain (Kelly et al., 2002).
The expression of α1 subunit mRNA in hippocampal dentate gyrus increased during postnatal development. Similar findings were reported in the past using in situ hybridization (Laurie et al., 1992), and single-cell RT-PCR from acutely isolated dentate granule cells (Brooks-Kayal et al., 2001). The immunoreactivity for α1 subunit polypeptide also increases during postnatal development of hippocampal dentate gyrus (Fritschy et al., 1994). The expression of α1 subunit increases during postnatal development in various other regions of the brain, such as neocortex (Poulter et al., 1992), superior colliculus (Juttner et al., 2001), substantia nigra (Veliskova et al., 1998), dorsolateral thalamus (Okada et al., 2000), cerebellar granule cells (Meinecke and Rakic, 1990; Tia et al., 1996) and recently in cerebellar stellate neurons (Vicini et al., 2001).
In contrast to the previous reports, we used high-resolution CLSM to describe the distribution of α1-subunit polypeptide in neonatal and adult rats, which clearly demonstrated punctate nature of α1 immunoreactivity. Clusters of α1 immunoreactivity were present outlining the cell bodies and dendrites of granule cells. The distribution of GABAA receptors on the neuronal membrane is typically non-uniform, with sites of high receptor densities at postsynaptic membranes of GABAergic synapses (Killisch et al., 1991; Fritschy et al., 1992; Craig et al., 1994; Nusser et al., 1995). The α1 subunit is known to cluster at synapses in vitro and in vivo (Wisden et al., 1992; Persohn et al., 1992). In contrast to adult hippocampus, few clusters of α1 immunreactivity were present in cell bodies of dentate granule cells of neonantal rats. One possible explanation for this finding is that there are only a few GABAergic synapses on immature dentate granule cells. This interpretation is supported by electrophysiological studies (Hollrigel and Soltesz, 1997), which show low frequency of IPSCs in dentate granule cells of neonatal rats, compared with that in granule cells of adult rats. Alternate possibility is that the GABAergic synapses are present on granule cells of neonatal rats but lack the α1 subunit.
Developmental changes in allopregnanolone sensitivity of GABAA receptors on DGCs may be a consequence of increased expression of the α1 subunit. Studies on native and recombinant GABAA receptors suggest that there is strong correlation between the allopregnanolone potentiation of GABAA receptor currents and type of α subunits being expressed. Increased expression of α2 and suppression of α1 subunit in magnocellular oxytocin neurons at the time of parturition diminishes allopregnanolone potentiation of GABA currents (Brussaard et al., 1997). ln recombinant GABAA receptors, allopregnanolone potentiation of recombinant GABA receptors is increased in the presence of α1, compared with that in the presence of α2 subunit (Shingai et al., 1991). However, other authors reported reduced sensitivity to allopregnanolone in the presence of α1 subunit in recombinant GABAA receptor (Maitra and Reynolds, 1999) and even inhibition of GABA current (Hauser et al., 1996). The δ subunit also alters neurosteroid modulation of GABAA receptors but the nature of this modulation remains controversial. Increased expression of δ subunit diminished THDOC sensitivity of GABAA receptors (Zhu et al., 1996a). In contrast, in δ subunit knockout mice, pregnanolone-induced sleep time was significantly reduced (Mihalek et al., 1999).
NINDS grants KO2-NS 02081 and RO1 NS40337 to J.K., and EF grant to Z.M. supported this work.