help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Heger, S.
Right arrow Articles by Tobet, S. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Heger, S.
Right arrow Articles by Tobet, S. A.
Endocrinology Vol. 144, No. 6 2566-2579
Copyright © 2003 by The Endocrine Society

Overexpression of Glutamic Acid Decarboxylase-67 (GAD-67) in Gonadotropin-Releasing Hormone Neurons Disrupts Migratory Fate and Female Reproductive Function in Mice

Sabine Heger, Marianne Seney, Elizabeth Bless, Gerald A. Schwarting, Marie Bilger, Alison Mungenast, Sergio R. Ojeda and Stuart A. Tobet

Division of Neuroscience, Oregon National Regional Primate Research Center, Oregon Health & Science University (S.H., M.B., A.M., S.R.O.), Beaverton, Oregon 97006; and The Shriver Center, University of Massachusetts Medical School (M.S., E.B., G.A.S., S.A.T.), Waltham, Massachusetts 02452

Address all correspondence and requests for reprints to: Stuart A. Tobet, Ph.D., Colorado State University, Department of Biomedical Sciences, 1680 Campus Delivery, Fort Collins, Colorado 80523. E-mail: stuart.tobet{at}colostate.edu; or Sergio R. Ojeda, D.V.M., Division of Neuroscience, Oregon National Regional Primate Research Center, Oregon Health & Science University, 505 NW 185th Avenue, Beaverton, Oregon 97006. E-mail: ojedas{at}ohsu.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
{gamma}-Aminobutyric acid (GABA) inhibits the embryonic migration of GnRH neurons and regulates hypothalamic GnRH release. A subset of GnRH neurons expresses GABA along their migratory route in the nasal compartment before entering the brain, suggesting that GABA produced by GnRH neurons may help regulate the migratory process. To examine this hypothesis and the possibility that persistence of GABA production by GnRH neurons may affect subsequent reproductive function, we generated transgenic mice in which the expression of glutamic acid decarboxylase-67 (GAD-67), a key enzyme in GABA synthesis, is targeted to GnRH neurons under the control of the GnRH gene promoter. On embryonic d 15, when GnRH neurons are still migrating, the transgenic animals had more GnRH neurons in aberrant locations in the cerebral cortex and fewer neurons reaching the hypothalamic-preoptic region, whereas migration into the brain was not affected. Hypothalamic GnRH content in mutant mice was low during the first week of postnatal life, increasing to normal values during infantile development (second week after birth) in the presence of increased pulsatile GnRH release. Consistent with these changes, serum LH and FSH levels were also elevated. Gonadotropin release returned to normal values by the time steroid negative feedback became established (fourth week of life). Ovariectomy at this time demonstrated an enhanced gonadotropin response in transgenic animals. Although the onset of puberty, as assessed by the age at vaginal opening and first ovulation, was not affected in the mutant mice, estrous cyclicity and adult reproductive capacity were disrupted. Mutant mice had reduced litter sizes, increased time intervals between deliveries of litters, and a shorter reproductive life span. Thus, GABA produced within GnRH neurons does not delay GnRH neuronal migration, but instead serves as a developmental cue that increases the positional diversity of these neurons within the basal forebrain. In addition, the results suggest that the timely termination of GABA production within the GnRH neuronal network is a prerequisite for normal reproductive function. The possibility arises that similar abnormalities in GABA homeostasis may contribute to syndromes of hypothalamic amenorrhea/oligomenorrhea in humans.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
MAMMALIAN puberty and mature female reproductive function require the tight regulation of a subset of highly specialized neurons that are located mostly in the preoptic region of the basal forebrain and secrete the neuropeptide GnRH (1, 2, 3). GnRH neuronal function is regulated by both synaptic inputs and glia to neuron signaling pathways (4, 5, 6). Although the latter are mostly excitatory (7), the neuronal control of GnRH neurons has both stimulatory and inhibitory components (reviewed in Ref. 8). The main inhibitory neurotransmitter in the central nervous system (CNS) is {gamma}-amino butyric acid (GABA). Almost 50% of all synaptic contacts in the hypothalamus are GABAergic (9). GABA exerts its actions on GnRH secretion via two membrane-anchored receptors, the GABAA receptor, a ligand-gated chloride channel, and the GABAB receptor, a seven-transmembrane domain receptor coupled to a G protein.

Most of the known actions of GABA on GnRH neurons appear to involve activation of GABAA receptors and have been shown to either inhibit or stimulate GnRH neuronal function. Experiments in nonhuman primates have best demonstrated that GABA, acting via these receptors, restrains GnRH release during female sexual development (10) and that the removal of this restraining influence results in increased GnRH secretion (11, 12, 13), thereby advancing the onset of puberty (13). In vivo studies in rats are, in general, consistent with this concept. For instance, GABA release in the preoptic area (POA) decreases before the LH discharge (14), and blockage of GABAA receptors during the afternoon of proestrus advances the time of the preovulatory LH surge (15). In contrast, intrahypothalamic infusion of GABA during this time abolishes the LH surge (16, 17). In both male and female rats, this inhibitory effect of the GABAergic system on GnRH secretion becomes established during sexual maturation (18, 19). The relevance of a GABA-dependent inhibitory control to the transsynaptic regulation of GnRH in humans was suggested by the ability of a GABA agonist to arrest the progression of puberty in a patient with central sexual precocity (20) associated with nonketonic hyperglycemia. This metabolic defect prevents the metabolism of glycine, an amino acid that cooperates with glutamate in the activation of N-methyl-D-aspartate receptors.

Seemingly contradicting these findings, other studies showed that infusion of GABA into the third ventricle elicited LH release (21, 22) and that exposure of rat median eminence fragments to GABA evokes GnRH secretion (23, 24). Subsequent studies using the GnRH cell line GT1-7 demonstrated that GABA acting via GABAA receptors excites these neurons (25) and stimulates GnRH release (26). Like GT1-7 cells and most brain neurons during fetal and early postnatal development (27), embryonic GnRH neurons respond with excitation to GABAA receptor activation (28). This property does not appear to be restricted to immature neurons, as electrophysiological evidence has been recently presented demonstrating that the majority of adult GnRH neurons in situ also respond to GABAA receptor activation with excitation (29). It would thus appear that throughout their natural history GnRH neurons are subjected to both GABA inhibitory and excitatory influences mediated by GABAA receptors. Although the former influences may require neuronal circuits functionally connected to the GnRH neuronal network, the latter is directly exerted on GnRH neurons.

GnRH neurons originate outside the CNS, in the olfactory placode, and migrate during embryogenesis across the nasal septum and cribriform plate until they reach their final destination within the forebrain (30, 31). During early development, a subset of embryonic GnRH neurons has the intrinsic capacity to produce GABA (32), raising the possibility that GABA produced within the GnRH neuronal network may contribute to regulating specific aspects of GnRH neuronal migration. Because GABAA receptor agonists delay GnRH neuronal migration (33, 34), it is possible that at least part of this regulatory influence reflects the existence of an internal, GABA-dependent regulatory loop used by GnRH neurons to coordinate their migratory behavior. Once GnRH neurons enter the brain, GABA production ceases, suggesting that this endogenous source of GABA is no longer required for the migration of GnRH neurons within the brain and/or that termination of GABA synthesis within the GnRH neuronal network is required for the normal secretion of GnRH postnatally.

The purpose of this study was to investigate the importance that GABA produced within the GnRH neuronal network may have in the control of GnRH neuronal migration and to understand the impact that persistence of GABA production within the GnRH neuronal network may have on GnRH neuronal function during postnatal life. To address these issues, we used a gain of function approach by which we targeted the gene encoding glutamic acid decarboxylase-67 (GAD-67), a key enzyme in GABA synthesis, to GnRH neurons using a promoter recently shown to accurately target reporter genes to these cells (35). The results show that GABA overproduction by GnRH neurons increases neuronal mistargeting within the brain, disrupts adult reproductive cyclicity, and results in premature central reproductive aging, probably because of alterations in pulsatile GnRH secretion. Preliminary reports of these findings have appeared (36, 37).


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Generation of transgenic mice
To generate animals in which GABA production is targeted to GnRH neurons, we used a rat cDNA encoding GAD-67, one of the two key enzymes in GABA synthesis (38), and an abbreviated version of the rat GnRH promoter (35, 39), henceforth referred to as the GnRH enhancer/promoter (enh/prom). Both DNA fragments were cloned (see below) into the pBA expression vector (40). This vector, derived from the pBS-KS+ plasmid (Stratagene, La Jolla, CA), has a 650-bp ß-globin intron fragment (41, 42) preceded by a multiple cloning site and a downstream 630-bp fragment containing the polyadenylation signal of the human GH gene (henceforth referred to as hGH poly A) (40) preceded by another multiple cloning site (Fig. 1Go). A 1972-bp DNA fragment containing the entire coding region [nucleotides (nt) -42 to +1930] of rat GAD-67 mRNA (38, 43) (GenBank accession no. M34445) was provided by A. Tobin (University of California, Los Angeles, CA). The plasmid pBS-SK II containing this cDNA was first linearized with SacII and blunted. Thereafter, the GAD-67 cDNA was excised with HindIII and inserted into the SmaI-HindIII sites of the pBA vector, between the ß-globin intron and the hGH poly A fragment, to generate pBA-GAD-67 (Fig. 1Go).



View larger version (18K):
[in this window]
[in a new window]
  		Figure 1. Diagram of the rat GnRH promoter/GAD-67/pBA transgene used to direct GAD-67 expression to GnRH neurons in transgenic mice. The transgene was excised from the pBA plasmid by digestion with BciVI and AgeI (bold letters). For a detailed description of the transgene, see the text.

 
The rat GnRH gene promoter (39, 44) we used is the fusion product of an approximately 300-bp enhancer region contained in the 5'-regulatory region of the rat gene (-1863 to -1571) and the proximal promoter region (-173 to +112) containing the GnRH gene transcription initiation site (Fig. 1Go). This fusion product directs high levels of reporter gene expression in both GT1-7 cells and transgenic mice, and accurately targets reporter gene expression to GnRH neurons in transgenic mice (35, 39). A recent study showed that this fusion product consistently provides an equal or higher degree of cell-specific expression to a reporter gene than the full-length mouse GnRH promoter (35).

The GnRH enh/prom cloned into the expression vector pGL-3 (provided by P. Mellon, University of California-San Diego, La Jolla, CA) was removed from pGL-3 by MluI and BglII digestion, blunted, and cloned 5' to the ß-globin intron fragment of pBA-GAD-67 into the blunted EagI site (Fig. 1Go). The resulting construct of approximately 4530 bp was excised with BciVI and AgeI (Fig. 1Go), releasing a fragment consisting of the GnRH enh/prom gene, the ß-globin intron, the rat GAD-67 cDNA, and the hGH poly A fragment. This DNA was then used for the production of transgenic mice. Transgenic mice were generated in a B6D2 background by the Oregon Health & Science University Transgenic Animal Facility, using standard procedures (45). Animals carrying the transgene were identified by PCR analysis of genomic DNA. The DNA was extracted (46) from tail or toe clips and amplified in a 25-µl PCR with the sense primer 5'-ACAGATAGACCAGCAGGTGTT-3' corresponding to the region -62 and -42 in the short GnRH promoter and the antisense primer 5'-GTCCATGGTGATACAAGGGACA-3' complementary to nt 819–845 in the original ß-globin intron sequence (41) (GenBank accession no. V00882), generating a PCR product of 355 bp (primers purchased from Invitrogen/Life Technologies, Inc., Carlsbad, CA). After analysis of the PCR products by agarose electrophoresis, two founders were identified. They were crossed with B6D2 wild-type mice to determine their ability to transmit the transgene to their progenies in a Mendelian fashion, and two transgenic lines (2210 and 2369) were selected for further study. Both lines were independently bred to homozygocity. The homozygous condition was confirmed by genotyping the progeny of presumptive homozygote mice backcrossed to B6D2 wild-type mice. Animals producing at least three consecutive litters in which all pups proved positive for the transgene were considered to be homozygous.

Transgenic and control B6D2 mice (The Jackson Laboratory, Bar Harbor, ME) were housed under controlled photoperiod (12 h of light, 12 h of darkness; lights on at 0700) and temperature (23–25 C), and were given free access to tap water and rodent chow. They were used in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals. The experimental protocols were approved by the Oregon National Regional Primate Research Center institutional research animal committee.

Assessment of transgene expression
To verify that the GAD-67 transgene is indeed functionally expressed in GnRH neurons, we performed double-immunolabeling experiments to detect GABA in embryonic GnRH neurons. Three developmental windows were selected for study: early in the process of neuronal migration [embryonic d 13 (E13)], at the three quarter point of the migratory process across the cribriform plate (E15), and at the end of the migratory process [postnatal d 0 (P0)]. Because these experiments revealed that mice of line 2210, but not those of line 2369, had an increased number of GnRH neurons containing GABA immunoreactivity, we performed a real-time RT-PCR (47) study in juvenile mice to verify that after GnRH neurons have settled into their final destination in the basal forebrain, transgene expression remains higher in the POA of line 2210 mice compared with line 2369.

Immunohistochemistry
Fetuses on E13 and E15 (plug on d 0) were delivered via cesarean section from isoflurane-anesthetized, timed pregnant mice. The embryos or newborns (P0) were anesthetized by keeping them on ice for 3–5 min and then they were transcardially perfused with either 4% paraformaldehyde-0.2% glutaraldehyde in 0.1 M phosphate buffer (PB; pH 7.4) or with 2% acrolein (Sigma-Aldrich Corp., St. Louis, MO) in 0.1 M PB (48, 49) using a hand-held syringe with a 30-gauge needle under a dissecting microscope. Heads and brains were postfixed in the same fixative for 6–24 h at 4 C and then transferred to 0.1 M PB.

Heads from wild-type or transgenic mouse embryos and newborns were embedded in 5% agarose and cut at 60 µm in a parasagittal plane using a vibrating microtome (VT1000S, Leica Corp., Deerfield, IL). Immunocytochemical procedures were similar to those previously reported (34). The GnRH decapeptide was detected using two different rabbit antisera: GF-6 (provided by N. Sherwood) or LR-1 (provided by Dr. R. Benoit). Immunoreactive GABA was detected in cells using a rabbit antiserum obtained from DiaSorin, Inc. (Stillwater, MN). To detect GABA in immunofluorescent double-labeling experiments, a guinea pig antiserum (Chemicon, Temecula, CA) was used at 1:200 dilution in conjunction with LR-1 (1:2,500 in paraformaldehyde/glutaraldehyde-fixed tissue) or at 1:500 dilution in conjunction with GF-6 (1:2,500 in acrolein-fixed tissue). For brightfield analyses, LR-1 was used in paraformaldehyde/glutaraldehyde-fixed tissue at 1:10,000 dilution, and GF-6 was used in acrolein-fixed tissue at 1:10,000 dilution, whereas GABA was detected using the DiaSorin, Inc., antiserum at 1:10,000 dilution in acrolein-fixed tissue. Antisera were diluted in 1.0% BSA in 0.05 M PBS with 0.3% Triton X-100 (pH 7.5). Tissue sections were pretreated with 0.1 M glycine in PBS (30 min)/0.5% sodium borohydride in PBS (15 min) and for at least 30 min in 5% normal goat serum (NGS) with 1% H2O2 and 0.3% Triton X-100 in PBS. Tissue sections were incubated with primary antisera over 2 or 3 nights with shaking at 4 C. After this incubation, sections for brightfield analyses were washed with PBS containing 1.0% NGS and 0.02% Triton X-100 at room temperature and then incubated with goat antirabbit IgG biotinylated secondary antibodies (Vector Laboratories, Inc., Burlingame, CA; 1:250) in NGS/PBS for 2 h. After secondary antisera, sections were washed with 0.05 M PBS and 0.02% Triton X-100 before incubation with Vectastain avidin-biotin peroxidase complex reagent for 1 h (Vector Laboratories, Inc.). Finally, sections were washed with 0.05 M Tris-buffered saline (TBS) before a dark gray/black reaction product was produced by horseradish peroxidase using 0.025% 3,3'-diaminobenzidine with 0.2% nickel ammonium sulfate in TBS as substrate with 0.02% hydrogen peroxide for 5 min. Sections were then washed with TBS, mounted on slides, and coverslipped using Permount (Fisher Scientific, Pittsburgh, PA). For fluorescent double labeling, a goat antirabbit conjugated with fluorescein isothiocyanate (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) was used to detect anti-GnRH primary antisera, and a donkey antiguinea pig conjugated with biotin (Jackson ImmunoResearch Laboratories, Inc.), followed by avidin Cy3 (Jackson ImmunoResearch Laboratories, Inc.), was used to detect the anti-GABA primary antiserum.

Real-time PCR
Details of the procedure in our hands have been published previously (50). The primers used for amplification were a sense forward primer (5'-CATCTCACGTAAGCTTGACTGACT-3') corresponding to a region that comprises the 3' end of GAD-67 cDNA (nt 2104–2119 in rat GAD-67 mRNA; GenBank accession no. M3445) plus an additional 8 bp of the pBSK backbone vector from which the pBA vector is derived (40), and a reverse primer (5'-GGGCCAGGAGAGGCACT-3'), complementary to a region located between nt 38 and 54 upstream from the consensus polyadenylation signal AATAA in hGH poly A. The internal fluorescent oligonucleotide probe used (5'-TGGCATCCCTGTGACCCCTCC-3'; PE Applied Biosystems, Foster City, CA) corresponds to a sequence located between nt 57–77 upstream of AATAA in hGH poly A. All three primers were selected with the assistance of the program Primers Express (PE Applied Biosystems) provided by Integrated DNA Technology Co. (Santa Clara, CA). A set of 18S ribosomal primers and an 18S fluorescent probe were used as an internal control to correct for procedural variability. These primers were purchased as a kit (TaqMan ribosomal RNA control reagents kit) from PE Applied Biosystems.

Measurement of hypothalamic GnRH content
To measure hypothalamic GnRH content, the POA and medial basal hypothalamus (MBH) of 4-, 8-, 12-, 16-, 20-, 24-, and 28-d-old transgenic mice and wild-type animals were dissected as described previously (51). The fragments were homogenized in 100 µl ice-cold 0.1 N HCl using a glass microhomogenizer, followed by trituration through long flat loading pipette tips. The homogenate was centrifuged for 15 min at 13,000 x g at 4 C, the supernatant was recovered, and the procedure was repeated to extract GnRH remaining in the pellet. Samples were stored at -20 C until assayed for GnRH. GnRH was assayed by an RIA procedure previously described (52) using [125I]GnRH as tracer and a rabbit polyclonal antiserum (HU60) that recognizes the fully processed, mature decapeptide (53). The antiserum was used at a 1:25,000 dilution; the assay had a sensitivity of 0.4 pg/tube.

Measurement of pulsatile hypothalamic GnRH release
To examine potential changes in pulsatile GnRH release that may occur in the transgenic mice, we used an in vitro system similar to that described by Bourguignon et al. (54). Because LH secretion was increased in transgenic mice by the end of the second week of postnatal life, we used hypothalamic fragments from 10- to 12-d-old wild-type and GAD-67 transgenic female mice. The fragments were dissected to include both the MBH and POA, and were incubated in individual flasks containing 250 µl Krebs-Ringer bicarbonate buffer at 37 C for 6 h under an atmosphere of 95 O2 and 5% CO2, as previously described (52), except that D-dextrose was used at 4.5 mg/ml instead of 1 mg/ml. After a preincubation period of 30 min, the medium from each flask was collected at 7.5-min intervals (54) and replaced with fresh medium. GnRH was measured in 150-µl aliquots. All samples were measured in a single assay to avoid interassay variability.

Measurement of serum gonadotropin levels
To measure LH and FSH serum levels, trunk blood was collected from the same animals used to determine hypothalamic GnRH content. LH and FSH levels were measured by RIAs as previously reported (55).

Evaluation of estrous cyclicity and reproductive competence
Female transgenic and wild-type animals were weaned on d 22 and housed as groups of four animals per cage. They were inspected daily for vaginal opening, and vaginal lavages were performed daily once vaginal opening had occurred. At 50 d of age, the transgenic females were mated with transgenic males of the same age (10 cages total, 1 male/1 female), and the occurrence of the first litter, litter size, and litter weights were recorded. To determine the effect of GAD-67 expression in GnRH neurons on the reproductive life span of mice, five transgenic and five wild-type breeding pairs were bred until the transgenic mice were no longer able to reproduce. In other experiments five transgenic females were individually bred to wild-type males to determine whether the lengthened mating-delivery interval (MDI), reduced number of pups per litter, and shortened reproductive life span previously observed in the transgenic females result from a female-specific deficiency or are due to a male defect.

Ovariectomy
Ovariectomy was performed on postnatal d 24. The animals were anesthetized with isoflurane, and the ovaries were aseptically removed via a single, dorsal skin incision, followed by blunt separation of the underlying muscle-aponeurosis interface. Different groups of animals were killed 1, 2, and 4 d later, and trunk blood was collected for gonadotropin assay.

Data analysis
Embryonic GnRH neurons.
Neurons containing immunoreactive GnRH in GAD-67 transgenic mice (E13, n = 8; E15, n = 9; P0, n = 7) and wild-type mice (E13, n = 9; E15, n = 7; P0, n = 6) were counted manually in three main compartments (nasal compartment, dorsal forebrain, and ventral forebrain) at x400 using an BH-2 microscope (Olympus Corp., New Hyde Park, NY). Cells were considered to be GnRH positive if they contained dense immunoreaction product, a standard size (10–15 µm diameter), and fusiform morphology. All GnRH counts were taken from sets of alternate sections, such that all numbers represent approximately one half of the GnRH cell population. The boundaries for the three main compartments were the same as previously reported (49). The cribriform plate provided the boundary between the nasal compartment and the dorsal forebrain. A line connecting the caudal-most point of the cortex and the rhinencephalic sulcus, found below the olfactory bulb, designated the boundary between dorsal and ventral forebrain. As a subgrouping of the dorsal forebrain, the number of GnRH neurons in cortical regions was also counted. Cells in the cortex were designated as those cells not found within or ventral to the olfactory bulb. Within the ventral forebrain, a subpopulation of cells caudal to the optic chiasm was also counted for a further delineation of those cells that should have migrated the farthest. Data are presented as the mean ± SEM. P < 0.05 was considered statistically significant.

Hormone content and reproductive parameters.
The differences between several groups were analyzed by ANOVA, followed by a Student-Newman-Keuls multiple comparison test for unequal replications. In those cases where the data showed a significant deviation from a normal distribution, Kruskal-Wallis one-way ANOVA on ranks was performed as a nonparametric test. The differences between two groups were analyzed by t test. In cases where the data did not show normal distribution, a Mann-Whitney rank-sum test was performed. Data are presented as the mean ± SEM. P < 0.05 was considered statistically significant.

Pulsatile GnRH release.
In vitro GnRH secretory profiles were analyzed using the PULSAR program (56). The depth criterion for splitting peaks was set at 2.5, and the cut-off parameters (Gn) for pulse identification, G1 to G5, were 4.01, 3.01, 2.0, 1.50, and 1.01, respectively. These parameters are similar to those providing a false positive error rate of 1% (56). Differences in the number of peaks, pulse amplitude, and pulse frequency between wild-type and transgenic mice were analyzed by t test. Differences in the frequency of pulses at various interpeak intervals were analyzed using the {chi}2 test for frequency distributions.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
More GnRH neurons show GABA expression in GAD-67 transgenic mice
Cells containing immunoreactive GABA were found throughout the forebrain and in distinctive cells in the nasal compartment of Swild-type and transgenic mice. In E13 animals, double-label fluorescence immunohistochemistry revealed that, in agreement with previous results in mice (32), GABA and GnRH were coexpressed in some cells within the posterior nasal compartment/cribriform plate region (Fig. 2Go, red and green, respectively; examples of double-stained neurons denoted by white arrows). In the GAD-67 transgenic mice, there were notably more GnRH and GABA double-labeled cells than in wild-type controls (Fig. 2Go, A vs. B). Although not all GnRH neurons in the nasal compartment contained immunoreactive GABA, all cells containing immunoreactive GABA in this region were also GnRH immunoreactive. This increased number of double-labeled neurons was observed in mice of the 2210 line, but not in 2369 animals, suggesting that the GnRH enh/prom-GAD-67 transgene was expressed at functionally significant levels only in 2210 mice. Real-time PCR analysis of RNA extracted from the POA of juvenile (28-d-old) transgenic and wild-type mice revealed that this was indeed the case. The POA of mice from line 2210 had 54.8 ± 15.3 (n = 4) arbitrary units of GAD-67 transgene/pg total RNA input vs. 0.11 ± 0.02 (n = 4) arbitrary units detected in mice from line 2369. Because of these results and those of the double-immunolabeling study, all subsequent experiments were performed using line 2210 only.



View larger version (39K):
[in this window]
[in a new window]
  		Figure 2. Images of sections processed for GnRH immunoreactivity (green) and GABA immunoreactivity (red) show neurons approaching and crossing the cribriform plate on E13. In transgenic mice (B), there are many more GABA-immunoreactive GnRH neurons than in wild-type animals (A; arrows). The scale bar in B represents 25 µm for both panels. In this and subsequent figures: WT, wild-type mice; GAD-67, transgenic mice expressing GAD-67 in GnRH neurons.

 
To obtain a quantitative estimate of this intranasal increase in GABA-positive cells, we counted the number of GABA-immunoreactive cells in sections of the nasal compartment processed for single-label GABA using 3,3'-diaminobenzidine/nickel reaction product in wild-type and 2210 transgenic animals on E13 and E15 (Fig. 3Go). Based on the double-label analysis these immunoreactive cells, which had the characteristics of GnRH neurons, were probably GnRH neurons. There were significantly (P < 0.05) more GABA-expressing cells in the nasal compartment of transgenic animals (120 ± 22; n = 4) compared with wild-type animals (24 ± 4; n = 3). This difference remained evident through E15 when there were 84 ± 14 (n = 3) GABA neurons in the nasal compartment of transgenic mice vs. 23 ± 2 (n = 2) in wild-type controls (Fig. 3CGo). The images shown in Fig. 3AGo (wild-type mice) and Fig. 3BGo (GAD-67 transgenic mice) illustrate these differences on E15.



View larger version (48K):
[in this window]
[in a new window]
  		Figure 3. GABA-expressing neurons in the nasal compartment and near the cribriform plate in sagittal sections on E15. A, Image from a WT animal showing few migrating cells containing immunoreactive GABA in the nasal compartment (small arrows). B, Image from a GAD-67 animal exhibiting many more cells containing immunoreactive GABA in the nasal compartment. The large arrow points to a grouping of GABAergic cells near the cribriform plate. The scale bars represent 100 µm in each panel. C, Number of migrating GABA-immunoreactive neurons in the nasal compartment of E13 and E15 wild-type and GAD-67 mice. In this and subsequent figures, bars are means, and vertical lines represent the SEM. The number of animals in each group was four GAD-67 transgenic and three WT on E13, and three GAD-67 transgenic and two WT on E15. *, P < 0.05; +, not evaluated statistically.

 
GnRH neuron migratory fate is altered in GAD-67 transgenic mice
GnRH neurons, as visualized with GF-6 or LR-1 antisera, were found in their normal migratory positions on E13, E15, and P0 within the nasal compartment and brain. To assess the differences in the final destination of GnRH neurons in the transgenic compared with the wild-type mice, neurons were counted within specific regions. From E13 to E15 to P0, the percentage of GnRH neurons in the nasal compartment declined from 40% to 17% to 7%, respectively, as the majority of GnRH neurons took up residence in the brain. On E13, there were no statistically significant differences in the distribution of GnRH neurons between the transgenic and wild-type embryos (data not shown). By E15, differences that were small trends on E13 became significant between wild-type and transgenic animals (Fig. 4Go). In the transgenic embryos on E15, there were significantly (P < 0.05) fewer GnRH neurons that had migrated into the ventral forebrain (Fig. 4DGo) and fewer still that had reached locations caudal to the optic chiasm (P < 0.01; Fig. 4FGo). Concomitantly, there were significantly more GnRH neurons found in the cortical compartment (P < 0.05; Fig. 4Go, B and E), perhaps as a consequence of not having turned toward the ventral forebrain. Images illustrating this difference in cerebro-cortical location of GnRH neurons are shown in Fig. 4AGo (wild-type mice) and Fig. 4BGo (transgenic mice). Even with these differences in cell location, the total number of GnRH neurons on E15 was not significantly different between wild-type (864 ± 55; n = 7) and transgenic (779 ± 48; n = 9; Fig. 4CGo) mice.



View larger version (94K):
[in this window]
[in a new window]
  		Figure 4. GnRH neurons detected with the GF-6 antibody in sagittal sections. A, WT animal. There are very few GnRH-positive cells in the cortex. B, GAD-67 animal. There are many more GnRH-positive cortical cells (denoted by arrows). The scale bars represent 100 µm in each panel. Counts of GnRH neurons show that the total numbers of cells in WT and GAD-67 mice were not significantly different on E15 (C). However, transgenic mice had significantly more GnRH neurons in the cortex (E) and significantly fewer cells ventrally in the basal forebrain (D). When the caudal-most subpopulation of cells was examined, the loss of cells in the transgenic was even more pronounced (F). The number of animals in each group was nine GAD-67 transgenic and seven WT. *, P < 0.05; **, P < 0.01 (vs. WT controls).

 
The images in Fig. 5Go, A and B, depict the normal locations of the most caudally located GnRH neuronal population from parasagittal and coronal perspectives, respectively, in wild-type mice on the day of birth (P0). In the transgenic animals on P0, there were again significantly (P < 0.05) fewer ventral and caudo-basal GnRH neurons than in wild-type mice (Fig. 5Go, D and E). Again the total cell number was not significantly different between wild-type (641 ± 38; n = 6) and transgenic (556 ± 40; n = 7; Fig. 5CGo) mice.



View larger version (107K):
[in this window]
[in a new window]
  		Figure 5. On the day of birth (P0), neurons containing immunoreactive GnRH (GF-6 antiserum) in MBH regions were located ventrally in WT mice. The caudal subpopulation was depleted in GAD-67 transgenic mice. A, Sagittal section, with arrows indicating those cells that were counted as part of the caudal subpopulation in relation to the optic tract in P0 mice. B, Coronal section indicating that caudal cells were most frequently located laterally, but occasionally medially in or around the arcuate nuclei in P0 mice. The scale bars represent 100 µm in each panel. Counts of GnRH neurons (C–E) show that the total numbers of cells in WT and GAD-67 transgenic mice were not significantly different on P0 (C). However, transgenic mice had significantly fewer cells ventrally in the basal forebrain (D). When the caudal-most subpopulation of cells was examined, the loss of cells in the transgenic was still significantly different (E). AC, Anterior commissure; CTX, cortex; LV, lateral ventricle; NC, nasal compartment; OB, olfactory bulb; OT, optic tract; 3V, third ventricle. The number of animals in each group was seven GAD-67 transgenic and six WT. *, P < 0.05.

 
GAD-67 overexpression in GnRH neurons reduces hypothalamic GnRH content during neonatal life
During the first week of postnatal life, the hypothalamic GnRH content was approximately 50% reduced in transgenic animals compared with control animals [4 d after birth, 62.6 ± 5.2 (n = 10) vs. 132.9 ± 14.7 pg/hypothalamus (n = 7; P = 0.001); 8 d, 54.6 ± 12.6 (n = 7) vs. 119.7 ± 10.7 pg/hypothalamus (n = 15; P < 0.01); Fig. 6AGo]. GnRH levels returned to normal values thereafter, suggesting an increased rate of GnRH synthesis (Fig. 6BGo).



View larger version (18K):
[in this window]
[in a new window]
  		Figure 6. Hypothalamic GnRH content in GAD-67 mice and WT control animals. A, During the first week of postnatal life, the hypothalamic GnRH content was reduced by approximately 50% in transgenic animals. *, P < 0.01. B, Thereafter, hypothalamic GnRH contents were similar in transgenic and control animals. In this and the next figure, circles represent means, and vertical lines are the SEM. Each group represents the mean of 4–15 mice.

 
GABA overexpression in GnRH neurons increases gonadotropin output during infantile development
During the infantile phase of rodent sexual development (second and third weeks of postnatal life), there is a transient activation of GnRH release (57) that results in increased pituitary gonadotropin hormone secretion. Although serum FSH levels are consistently elevated at this time (1), LH values are only sporadically increased (58), and thus are not easily detectable in single blood samples. These changes, originally described in rats, also occur in mice (59). GAD-67 transgenic mice showed a striking elevation in serum LH levels during the second and third weeks of postnatal life compared with wild-type controls (Fig. 7AGo; P < 0.01 on d 12 and P < 0.05 on d 16). By postnatal d 20, i.e. at the beginning of the juvenile period (1), serum LH levels decreased toward normal values (Fig. 7AGo). In both wild-type and transgenic mice, serum FSH levels rose during the first 2 wk after birth and reached maximum levels by d 12 (Fig. 7BGo). Thereafter, FSH levels decreased in wild-type mice, but remained significantly elevated in transgenic animals throughout the infantile period, i.e. during the third week of postnatal life (d 16, P < 0.01; d 20, P < 0.05). By d 24, levels were similar to those in wild-type mice (Fig. 7BGo).



View larger version (30K):
[in this window]
[in a new window]
  		Figure 7. Serum gonadotropin levels during the prepubertal development of WT and GAD-67 mice. Both serum LH (A) and FSH (B) were increased in transgenic mice during infantile development, but returned toward normal values during juvenile development at the time when ovarian steroid negative feedback became established. A1 and B1, Both the LH and FSH responses to removal of ovarian negative feedback were increased in GAD-67 mice compared with WT controls. Each group represents the mean of 7–18 mice. *, P < 0.05; **, P < 0.01.

 
These results suggested that GAD-67 overexpression in GnRH neurons enhances GnRH secretion during early postnatal development, and that establishment of ovarian steroid negative feedback during juvenile development counteracts this stimulatory effect, bringing gonadotropin secretion toward basal values. To experimentally test this idea, we ovariectomized mice on d 24 and determined serum LH and FSH levels 1, 2, and 4 d later. In both groups, ovariectomy led to a significant increase in LH and FSH (Fig. 7Go, A1 and B1). However, the increase in both LH and FSH, but particularly LH, was greater in genetically altered than in wild-type animals (P < 0.05; Fig. 7Go, A1 and B1), suggesting that GABA overproduction in GnRH neurons may facilitate GnRH release in response to the loss of steroid inhibition.

GABA overexpression in GnRH neurons accelerates pulsatile GnRH release in infantile mice
To determine whether the increased circulating LH levels seen in infantile mice are related to alterations in pulsatile GnRH release, we measured GnRH release from explants of hypothalamic tissue containing the POA and MBH regions. Detection of GnRH in medium samples collected every 7.5 min revealed that the POA-MBH of transgenic mice generated a significantly greater (P < 0.05) number of GnRH pulses in the 6-h period studied than wild-type controls (Fig. 8G1Go), resulting in a greater pulse frequency, as expressed in pulses per hour (Fig. 8HGo). In contrast, pulse amplitude was similar in both groups (Fig. 8G2Go). To determine whether there were differences in the length of the interpulse interval, we grouped the pulses into 3 categories: occurring at intervals of less than 30 min, between 31 and 60 min, or more than 60 min. Analysis of these data revealed that 21 of 36 pulses (56.76%) generated by transgenic POA-MBH fragments occurred at intervals shorter than 30 min, whereas only 6 of 19 pulses (31.58%) in wild-type mice occurred at this interval (P < 0.01, by {chi}2 test; Fig. 8IGo). The frequency of pulses occurring at intervals of 31–60 min was similar in both groups. In contrast, the POA-MBH of wild-type mice generated a significantly greater (P < 0.01) fraction of GnRH pulses occurring at intervals longer than 60 min than in transgenic animals [6 of 19 (31.57%) vs. 6 of 37 (16.22%) in transgenic mice]. Figure 8Go, A–C, depicts representative profiles of pulsatile GnRH release generated by the POA-MBH of wild-type mice. Figure 8Go, D–F, shows representative GnRH profiles of GAD-67 mutant mice.



View larger version (43K):
[in this window]
[in a new window]
  		Figure 8. The frequency, but not the amplitude, of pulsatile GnRH release is enhanced in GAD-67 mice. A–C, Representative profiles of pulsatile in vitro GnRH release from the POA-MBH of 12-d-old WT mice; D–F, representative profiles of GAD-67 mice. Arrows point to pulses of GnRH secretion identified by the PULSAR program. G1, The POA-MBH of GAD-67 mice shows a greater number of GnRH pulses in the 6-h period studied than WT controls; G2, the amplitudes of the pulses are similar in both groups; H, the frequency of GnRH pulses, expressed as pulses per hour, is significantly increased in GAD-67 mice compared with WT controls; I, the episodes of GnRH release occur at more regular intervals in the transgenic than in WT mice. *, P < 0.05; **, P < 0.01. The data shown were obtained from six WT and five GAD-67 mice.

 
Estrous cyclicity and reproductive function are disrupted in GAD-67 transgenic mice
The prepubertal profile of serum gonadotropin levels and the enhanced gonadotropin response to ovariectomy observed in the mutant mice at this time of development suggested that despite the facilitatory effect on GnRH release exerted by GABA expressed within the GnRH neuronal network, prepubertal gonadotropin secretion is kept in check by a strong negative feedback action of gonadal steroids. Consistent with this concept, the ages at vaginal opening and first estrous were similar in wild-type mice and transgenic animals [wild-type (n = 14) vs. 2210 (n = 12), 29.1 ± 1.0 vs. 30.1 ± 1.3 d and 41.2 ± 0.9 vs. 42.4 ± 1.6 d; Fig. 9Go, A and B]. In contrast to this lack of effect on the onset of puberty, estrous cyclicity was significantly disrupted in transgenic animals. These animals experienced longer periods of estrus than wild-type controls (44.6 ± 5.1% vs. 25.3 ± 2.8%; n = 12 for both groups; P < 0.01) and spent significantly less time in diestrus (31.1 ± 4.3% vs. 45.5 ± 4.7%; P < 0.05; n = 12 for both groups; Fig. 9CGo). In addition to clearly distinguishable estrous cycle phases, both wild-type and transgenic mice showed estrous-diestrous intermediate phases at a similar frequency (16.6 ± 2.6% vs. 11.3 ± 2.4%, respectively; Fig. 9CGo). Representative estrous cycle profiles demonstrating these differences are depicted in Fig. 9DGo. In addition to the aforementioned alterations in estrous cyclicity, reproductive function was significantly compromised in transgenic animals. The interval between exposing female mice to a male and delivery of the first litter (MDI) was lengthened (P < 0.05; Fig. 9EGo), and the number of pups delivered in the first litter was reduced (P < 0.05) in the mutant mice (Fig. 9FGo). This reduction, although subtle, remained evident in consecutive litters. The analysis of 43 litters in transgenic animals and 60 litters in wild-type animals showed that the transgenic mice delivered 6.0 ± 0.5 pups/liter compared with 7.5 ± 0.5 pups/litter in wild-type animals (P < 0.05; n = 5 breeding pairs/group). Birth weight was not altered in transgenic offspring [transgenic, 1.54 ± 0.03 g/pup (n = 37); wild-type, 1.59 ± 0.02 g/pup (n = 51)].



View larger version (47K):
[in this window]
[in a new window]
  		Figure 9. Reproductive function is disrupted in mice overexpressing GAD-67 in GnRH neurons. Transgenic and WT mice exhibit vaginal opening (A) and first estrus (B) at similar ages, demonstrating that the onset of puberty is not affected in the mutant animals. C, In contrast, estrous cyclicity is disrupted in transgenic mice, which exhibit almost twice as many days in estrous as control animals. D, Representative profiles illustrating the disruption of estrous cyclicity in GAD-67 mice compared with wild-type controls. Two animals per group are shown. E, The interval between exposing the female mice to a male and the delivery of the first litter (MDI) is lengthened in the mutant mice. F, The number of pups delivered in the first litter is reduced in transgenic animals. G, The mutant mice stop producing litters earlier than wild-type mice. H, The MDI increases with age in transgenic animals, but remains constant in wild-type controls for at least 12 consecutive litters. P, Proestrous; E, estrous; D, diestrous; ED, estrous-diestrous intermediate phase; D1, diestrous d 1; D2, diestrous d 2. *, P < 0.05; **, P < 0.01. Numbers above bars are the number of mice per group.

 
Because a lengthened MDI was observed in transgenic females bred to transgenic males, we performed additional experiments to determine whether this reproductive dysfunction is due to a female deficiency or to a combination of a reproductive defect in both sexes. Breeding GAD-67 transgenic female mice to wild-type males (n = 5 breeding pairs) demonstrated that the MDI of the transgenic females was significantly longer than that of wild-type animals [32.1 ± 3.4 d (n = 14 litters) vs. 22.9 ± 0.39 d (n = 60 litters) from wild-type animals; P < 0.01]. Thus, the expression of the GAD-67 transgene in males is not responsible for the decreased capacity of mutant mice to become pregnant in a timely fashion.

In addition to these deficiencies, the reproductive life span of the mutant female mice was shortened (Fig. 9GGo). The fraction of mutant females able to deliver litters persistently decreased with time, so that after delivering the ninth litter, 80% (four of five) of the transgenic animals were no longer able to reproduce (Fig. 9GGo). In contrast, wild-type mice remained fertile, delivering 12 or more litters before the experiment was discontinued (Fig. 9GGo). The premature reproductive aging observed in the mutant mice was also demonstrated by an increased time interval between delivery of litters (Fig. 9HGo). Wild-type mice delivered litters every 19–34 d, with an average MDI of 23 d. This interval remained constant regardless of whether the mice had delivered 1–3 or 9–12 litters (Fig. 9HGo). In contrast, the transgenic females delivered their litters at much more variable intervals (18–75 d) and showed a lengthening of these intervals as they grew older, so that after they produced their seventh litter, the interlitter interval had almost doubled (P < 0.05; Fig. 9HGo).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The present study demonstrates that increasing GABA production within the GnRH neuronal network, via the targeted expression of a GAD enzyme to GnRH neurons, alters the migratory fate of these cells and disrupts mature reproductive function in female mice.

GABA is synthesized from glutamate via a decarboxylation reaction mediated by GAD enzymes. In the CNS of vertebrates, these enzymes occur as two forms encoded by separate genes: a cytosolic product of 67 kDa and a membrane-targeted form of 65 kDa (60, 61). We chose to constitutively express the 67-kDa isoenzyme in GnRH neurons for several reasons: 1) GAD-67 is expressed earlier than GAD-65 in the developing brain (62); 2) it is expressed in neurons as an active holoenzyme, which allows it to rapidly catalyze GABA synthesis in the presence of glutamate (38); 3) gene targeting studies have demonstrated that GAD-67 is the most critical enzyme for GABA synthesis, as deletion of GAD-67 gene leads to more than 90% loss of brain GABA content (63), in contrast to the absence of GAD-65, which does not result in GABA depletion (64); 4) cells expressing the same GAD-67 cDNA used in the present study under the control of the tetracycline-dependent gene expression system respond with a rapid increase (within 1 h) in GABA synthesis when challenged with glutamate (65); and 5) when these genetically engineered cells are grafted near GnRH neurons, they increase GnRH secretion (as indicated by changes in gonadotropin output), indicating that endogenous glutamate is present at sufficiently elevated levels to be used as a substrate for GABA synthesis in cells expressing the GAD-67 enzyme.

In the present study, we show the presence of GABA in GnRH neurons of mice carrying the GnRH enh/prom-GAD-67 transgene and thus demonstrate that, as predicted by the observations mentioned above, the transgene was functional and resulted in high intracellular levels of GABA in these cells. In agreement with Lawson et al. (35), who reported no ectopic expression of a reporter gene under the transcriptional control of the GnRH enh/prom used in our study, we did not detect GABA overexpression outside the GnRH neuronal network. Although GnRH enh/prom has been demonstrated to specifically direct high levels of reporter gene expression to GnRH neurons in transgenic mice, it targets only about 25% of these cells (35) compared with more than 85% reported for the full 3 kb of the mouse 5'-regulatory region (66, 67, 68). Our results are consistent with these observations, as we detected GABA in a comparable fraction of GnRH neurons. Importantly, our results also demonstrate that targeting only a subpopulation of GnRH neurons is sufficient to alter the migratory fate of a subset of cells and has a significant impact on reproductive physiology.

Previous studies have shown that GABA inhibits GnRH neuronal migration via activation of GABAA receptors (33, 34). GABA is available to migrating GnRH neurons from two sources: one provided by fibers from cells in the ventral aspect of the olfactory pit and ventral olfactory epithelium (69), and a subset of embryonic GnRH neurons, which themselves have the capability to produce GABA (32). Our study shows that when GABA is produced by GnRH neurons, a subset of these cells migrates into aberrant locations in the cortex, and fewer neurons find their way to appropriate sites in the hypothalamus during embryonic development. At birth, few GnRH neurons are found in the cortex any longer, but the number of neurons in the hypothalamus is still reduced. This suggests that GnRH neurons in inappropriate locations are functionless and may either stop expressing GnRH (rendering them unidentifiable) or undergo apoptosis (70).

It has been previously shown that the GABAA receptor agonist muscimol inhibits GnRH neuronal migration in olfactory explants (33) and mice (34). More GnRH neurons remained in the nasal compartment in muscimol-treated animals, whereas bicuculline (a GABAA receptor antagonist) treatment led to disorganized GnRH cell distribution in the forebrain (34). A delay in neuronal migration similar to that caused by muscimol would have been expected in our experiments if GABA produced within GnRH neurons had been engaged in significant autocrine control. This was clearly not the case, as only the final destination of GnRH neurons was affected in GAD-67 mice. One possible explanation for this lack of effect on migration is that GABAA receptors on GnRH neurons that produce GABA are desensitized due to prolonged stimulation (29). Another possibility is that excess GABA may also be stimulating GABAB receptors on GnRH neurons (71). Although we have detected GABAB receptor expression in migrating GnRH neurons, we found no evidence that their activation could influence the migration of these cells (72). A third, and perhaps more tenable, explanation is that, in contrast to GABA produced by non-GnRH cells in the olfactory pit, synthesis of GABA within GnRH neurons is not a factor that influences the timely exit of these neurons from the nasal compartment.

Instead of delayed neuronal migration, GAD-67 mice had an increased number of GnRH neurons taking up residence in aberrant positions in the cerebral cortex and a loss of neurons reaching appropriate destinations in the basal forebrain. Because the migration of GnRH neurons is axophilic, these results suggest that overexpression of GABA by GnRH neurons interferes with some of the guiding mechanisms that target the neurons to specific locations in the brain. Although our study does not identify such mechanisms, some possible explanations can be entertained. GnRH neurons migrate in contact with the vomeronasal nerve (34, 73), and this relationship requires the guidance molecule netrin-1 and its receptor DCC (deleted in colorectal cancer). In the absence of DCC GnRH neurons migrate to inappropriate destinations (49), so that more GnRH neurons migrate to the cortex on E13 and E15 as they follow their aberrantly targeted vomeronasal guidance fibers (49). The migratory pattern in our transgenic animals could be viewed as a milder version of the defect observed in DCC-/- mice and may reflect the existence of hitherto unknown regulatory interactions between GABA and the netrin/DDC guidance system. Alternatively, an excess of GABA may have affected the migratory fate of some GnRH neurons by virtue of its ability to stimulate neuronal chemotaxis and chemokinesis during embryonic development (74) and to direct, via these mechanisms, the migration of neurons toward cortical regions of the brain (62). If, on the other hand, excess GABAergic stimulation generated by expression within GnRH neurons causes desensitization, as discussed above, then one mechanisms of aberrant migration in the current study may be the inappropriate dissociation of GnRH neurons from their guiding fibers. Although previous studies have already indicated that GABA may play a role in maintaining the relationship of GnRH neurons to their vomeronasal guidance fibers (34), a preliminary examination of peripherin immunoreactive fibers in the GAD-67 transgenic animals showed no evidence of fiber disruption (data not shown). Furthermore, we did not observe the same degree of disorganization of GnRH neurons as they entered the forebrain as we had observed after bicuculline treatment (34), indicating that desensitization is not the likely mechanism underlying the disruption of GnRH neuronal homing observed in GAD-67 animals. Further experimentation is required to clarify this issue.

In addition to the aforementioned targeting defects, GAD-67 mice exhibited significant alterations in adult reproductive function. They did not, however, show alterations in the time of puberty, as previously seen in rats in which GABA production was increased near GnRH nerve terminals after the grafting of cells genetically modified to produce GABA from endogenous glutamate (65). Although on their own, these findings suggest that GABA acting on or produced by GnRH neurons does not inhibit GnRH secretion in rodents, they also parenthetically argue once more against the possibility that the defects in neuronal targeting observed in GAD-67 mice are due to GABAA receptor desensitization. Should this have occurred, a delay in puberty would have been expected, because mice treated prenatally with bicuculline show a delay in vaginal opening in conjunction with a disruption in the location of GnRH neurons in the region of the organum vasculosum of the lamina terminalis (75).

In transgenic mice, the hypothalamic GnRH content was reduced during the first week of life. This might be related to the reduced number of GnRH neurons migrating into the caudal hypothalamus during development. The fact that the hypothalamic GnRH content returned to normal values after the first week of life indicates that the remaining GnRH neurons were able to compensate for the deficiency by increasing GnRH production. We also observed that at the time when GnRH content was increasing to normal values, circulating LH and FSH were elevated as well. These two observations and the unexpected finding that the frequency of pulsatile GnRH release was increased in transgenic mice by the end of the second week of life (when plasma gonadotropin levels were maximally elevated) indicate that GABA production within the GnRH network actually stimulates GnRH neurons to secrete more GnRH. Although such a stimulatory effect had been previously shown by both in vivo and in vitro studies (21, 22, 23, 24, 25, 26), it is only recently that a direct GABAA receptor-mediated excitatory effect of GABA on GnRH neurons was unambiguously shown to be a prevailing mode of GABA action on these cells throughout postnatal life (29).

Although both GnRH and gonadotropin output were increased in the mutant animals during infantile development, circulating gonadotropin levels returned to control values by the third week of life, coinciding with the establishment of a functional steroid negative feedback mechanism (1). This finding suggests that the stimulatory effect of GABA overproduction within the GnRH neuronal network, detected in the absence of steroid feedback inhibition, is efficiently counteracted by the negative feedback ovarian steroids and explains the normal timing of puberty observed in transgenic animals. The existence of such a feedback inhibition opposing the stimulatory effect of GABA was most clearly shown by the enhanced release of LH and FSH observed in transgenic mice ovariectomized at the end of juvenile development, i.e. at the time when negative steroid feedback inhibition is fully operative.

Although changes in prepubertal hormone release were well defined, the most striking alterations in reproductive function observed in GAD-67 mice occurred after puberty. Adult female mice showed a disrupted estrous cyclicity and an overall reduction in reproductive capacity, manifested as the generation of litters of reduced size, a lengthening of the intervals between deliveries, and the premature termination of reproductive competence. These alterations were remarkably similar to those previously described in rats carrying GABA-producing cells near GnRH nerve terminals (65) and were tantalizingly consistent with the recent observation that patients suffering from oligomenorrhea associated to polycystic ovarian syndrome and showing an abnormal pattern of pulsatile LH secretion also have increased GABA levels in cerebrospinal fluid (76). These observations and the peculiar pattern of GnRH/gonadotropin release observed in GAD-67 mice during prepubertal development suggest that the persistence of GABA production in GnRH neurons after they reach their final destination disrupts the normal interplay of central and peripheral regulatory inputs affecting GnRH secretion. According to this concept, GnRH neurons producing GABA would not only be subjected to a greater central excitatory input, but they would also escape more rapidly to small decreases in steroid negative feedback tone. As a result of this state of hyperexcitability, their overall responsiveness to modulatory influences would be compromised, resulting in undesirable increases in GnRH output. Such increases would, in turn, cause subtle, but destabilizing, increases in gonadotropin secretion able to disrupt the normal homeostasis of the estrous cycle. GABA produced in GnRH neurons can enhance GnRH secretion by at least three different mechanisms: 1) inhibiting inhibitory neuronal circuitries synaptically connected to GnRH neurons, 2) activating astrocytes in contact with GnRH neurons (77), and 3) directly activating GnRH secretion (29) via autocrine/paracrine mechanisms.

In summary, our results indicate that GABA produced by GnRH neurons influences the migratory fate of these cells by increasing their positional diversity at the end of the migratory process. In addition, the results suggest that the timely termination of GABA production within the GnRH neuronal network (as seen in normal animals when GnRH neurons enter the brain) is a prerequisite for normal reproductive function. Based on these findings, the possibility needs to be considered that an inappropriately increased GABA production by and/or near GnRH neurons may be a factor contributing to syndromes of hypothalamic amenorrhea/oligomenorrhea in humans.


   Acknowledgments
 
We thank A. Tobin for kindly providing us with the rat GAD-67 cDNA, P. L. Mellon for the generous gift of the rat GnRH promoter, and L. Dees for performing the RIAs for LH and FSH. We also thank M. E. Costa for her expert technical assistance, and L. Goodspeed for her excellent care of the mice.


   Footnotes
 
This work was supported in part by the NICHD/NIH through Cooperative Grant U54-HD-18185-16 (to S.R.O.) as part of the Specialized Cooperative Centers Program in Reproduction Research, a Pediatric Endocrinology Fellowship from the European Society for Pediatric Endocrinology (to S.H.), and NIH Grants RR-00163 for the support of the Oregon National Primate Research Center (to S.R.O.) and R01-HD-33441 (to S.A.T. and G.A.S.).

Abbreviations: CNS, Central nervous system; DCC, deleted in colorectal cancer; E, embryonic day; GABA, {gamma}-aminobutyric acid; GAD-67, glutamic acid decarboxylase-67; MBH, medial basal hypothalamus; MDI, mating-delivery interval; NGS, normal goat serum; nt, nucleotide; p, postnatal day; PB, phosphate buffer; POA, preoptic area; TBS, Tris-buffered saline.

Received October 25, 2002.

Accepted for publication February 6, 2003.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Ojeda SR, Urbanski HF 1994 Puberty in the rat. In: Knobil E, Neill JD, eds. The physiology of reproduction. 2nd ed. New York: Raven Press; vol 3:363–409
  2. Plant TM 1994 Puberty in primates. In: Knobil E, Neill J, eds. The physiology of reproduction. 2nd ed. New York: Raven Press; vol 2:453–485
  3. Terasawa E 1995 Mechanisms controlling the onset of puberty in primates: the role of GABAergic neurons. In: Plant TM, Lee PA, eds. The neurobiology of puberty. Bristol: Journal of Endocrinology Ltd.; 139–151
  4. Terasawa E, Fernandez DL 2001 Neurobiological mechanisms of the onset of puberty in primates. Endocr Rev 22:111–151[Abstract/Free Full Text]
  5. Brann DW, Mahesh VB 1997 Excitatory amino acids: evidence for a role in the control of reproduction and anterior pituitary hormone secretion. Endocr Rev 18:678–700[Abstract/Free Full Text]
  6. Ojeda SR, Ma YJ, Lee BJ, Prevot V 2000 Glia-to-neuron signaling and the neuroendocrine control of female puberty. Recent Prog Horm Res 55:197–224
  7. Ojeda SR, Prevot V, Heger S 2001 Regulation of puberty. Curr Opin Endocrinol Diabetes 8:154–160[CrossRef]
  8. Ojeda SR, Terasawa E 2002 Neuroendocrine regulation of puberty. In: Pfaff D, Arnold A, Etgen A, Fahrbach S, Moss R, Rubin R, eds. Hormones, brain and behavior. New York: Elsevier; vol 4:589–659
  9. Decavel C, van den Pol AN 1990 GABA: a dominant neurotransmitter in the hypothalamus. J Comp Neurol 302:1019–1037[CrossRef][Medline]
  10. Mitsushima D, Hei DL, Terasawa E 1994 {gamma}-Aminobutyric acid is an inhibitory neurotransmitter restricting the release of luteinizing hormone-releasing hormone before the onset of puberty. Proc Natl Acad Sci USA 91:395–399[Abstract/Free Full Text]
  11. Mitsushima D, Marzban F, Luchansky LL, Bruich AJ, Keen KL, Durning M, Golos TG, Terasawa E 1996 Role of glutamic acid decarboxylase in the prepubertal inhibition of the luteinizing hormone releasing hormone release in female rhesus monkeys. J Neurosci 16:2563–2573[Abstract/Free Full Text]
  12. Terasawa E, Luchansky LL, Kasuya E, Nyberg CL 1999 An increase in glutamate release follows a decrease in {gamma} aminobutyric acid and the pubertal increase in luteinizing hormone releasing hormone in female rhesus monkeys. J Neuroendocrinol 11:275–282[CrossRef][Medline]
  13. Keen KL, Burich AJ, Mitsushima D, Kasuya E, Terasawa E 1999 Effects of pulsatile infusion of the GABAA receptor blocker bicuculline on the onset of puberty in female rhesus monkeys. Endocrinology 140:5257–5266[Abstract/Free Full Text]
  14. Jarry H, Hirsch B, Leonhardt S, Wuttke W 1992 Amino acid neurotransmitter release in the preoptic area of rats during the positive feedback actions of estradiol on LH release. Neuroendocrinology 56:133–140[Medline]
  15. Mitsushima D, Jinnai K, Kimura F 1997 Possible role of the {gamma}-aminobutyric acid-A receptor system in the timing of the proestrous luteinizing hormone surge in rats. Endocrinology 138:1944–1948[Abstract/Free Full Text]
  16. Lamberts R, Vijayan E, Graf M, Mansky T, Wuttke W 1983 Involvement of preoptic-anterior hypothalamic GABA neurons in the regulation of pituitary LH and prolactin release. Exp Brain Res 52:356–362[Medline]
  17. Herbison AE, Dyer RG 1991 Effect on luteinizing hormone secretion of GABA receptor modulation in the medial preoptic area at the time of proestrous luteinizing hormone surge. Neuroendocrinology 53:317–320[Medline]
  18. Moguilevsky JA, Carbone S, Szwarcfarb B, Rondina D 1991 Sexual maturation modifies the GABAergic control of gonadotrophin secretion in female rats. Brain Res 563:12–16[CrossRef][Medline]
  19. Mitsushima D, Kimura F 1997 The maturation of GABAA receptor-mediated control of luteinizing hormone secretion in immature male rats. Brain Res 748:258–262[CrossRef][Medline]
  20. Bourguignon J-P, Jaeken J, Gerard A, de Zegher F 1997 Amino acid neurotransmission and initiation of puberty: evidence from nonketotic hyperglycinemia in a female infant and gonadotropin-releasing hormone secretion by rat hypothalamic explants. J Clin Endocrinol Metab 82:1899–1903[Abstract/Free Full Text]
  21. Ondo JG 1974 {gamma}-Aminobutyric acid effects on pituitary gonadotropin secretion. Science 186:738–739[Abstract/Free Full Text]
  22. Vijayan E, McCann SM 1978 The effects of intraventricular injection of {gamma}-aminobutyric acid (GABA) on prolactin and gonadotropin release in conscious female rats. Brain Res 155:35–43[CrossRef][Medline]
  23. Nikolarakis KE, Loeffler J-PH, Almeida OFX, Herz A 1988 Pre- and postsynaptic actions of GABA on the release of hypothalamic gonadotropin-releasing hormone (GnRH). Brain Res Bull 21:677–683[CrossRef][Medline]
  24. Masotto C, Wisniewski G, Negro-Vilar A 1989 Different {gamma}-aminobutyric acid receptor subtypes are involved in the regulation of opiate-dependent and independent luteinizing hormone-releasing hormone secretion. Endocrinology 125:548–553[Abstract]
  25. Hales TG, Sanderson MJ, Charles AC 1994 GABA has excitatory actions on GnRH-secreting immortalized hypothalamic (GT1-7) neurons. Neuroendocrinology 59:297–308[CrossRef][Medline]
  26. Martínez de la Escalera G, Choi ALH, Weiner RI 1994 Biphasic gabaergic regulation of GnRH secretion in GT1 cell lines. Neuroendocrinology 59:420–425[Medline]
  27. Cherubini E, Gaiarsa JL, Ben-Ari Y 1991 GABA: an excitatory transmitter in early postnatal life. Trends Neurosci 14:515–519[CrossRef][Medline]
  28. Kusano K, Fueshko S, Gainer H, Wray S 1995 Electrical and synaptic properties of embryonic luteinizing hormone-releasing hormone neurons in explant cultures. Proc Natl Acad Sci USA 92:3918–3922[Abstract/Free Full Text]
  29. DeFazio RA, Heger S, Ojeda SR, Moenter SM 2002 Activation of A-type {gamma}- aminobutyric acid receptors excites gonadotropin-releasing hormone neurons. Mol Endocrinol 16:2872–2891[Abstract/Free Full Text]
  30. Schwanzel-Fukuda M, Pfaff DW 1989 Origin of luteinizing hormone-releasing hormone neurons. Nature 338:161–164[CrossRef][Medline]
  31. Wray S, Grant P, Gainer H 1989 Evidence that cells expressing luteinizing hormone-releasing hormone mRNA in the mouse are derived from progenitor cells in the olfactory placode. Proc Natl Acad Sci USA 86:8132–8136[Abstract/Free Full Text]
  32. Tobet SA, Chickering TW, King JC, Stopa EG, Kim K, Kuo-Leblank V, Schwarting GA 1996 Expression of {gamma}-aminobutyric acid and gonadotropin-releasing hormone during neuronal migration through the olfactory system. Endocrinology 137:5415–5420[Abstract]
  33. Fueshko SM, Key S, Wray S 1998 GABA inhibits migration of luteinizing hormone-releasing hormone neurons in embryonic olfactory explants. J Neurosci 18:2560–2569[Abstract/Free Full Text]
  34. Bless EP, Westaway WA, Schwarting GA, Tobet SA 2000 Effects of {gamma}-aminobutyric acid(A) receptor manipulation on migrating gonadotropin- releasing hormone neurons through the entire migratory route in vivo and in vitro. Endocrinology 141:1254–1262[Abstract/Free Full Text]
  35. Lawson MA, MacConell LA, Kim J, Powl BT, Nelson SB, Mellon PL 2002 Neuron-specific expression in vivo by defined transcription regulatory elements of the GnRH gene. Endocrinology 143:1404–1412[Abstract/Free Full Text]
  36. Heger S, Bless E, Seney M, Bilger M, Ojeda S, Tobet S, Overexpression of glutamic acid decarboxylase-67 (GAD-67) in LHRH neurons disrupts neuronal migration and reproductive function in mice. Program of the 84th Annual Meeting of The Endocrine Society, San Francisco, CA, 2002, p 175 (Abstract P1-88)
  37. Heger S, Bless E, Seney M, Bilger M, Ojeda SR, Tobet SA, Glutamic acid decarboxylase-67 (GAD-67) overexpression in LHRH neurons alters neuronal migration, estrous cyclicity and reproduction in female mice. Program of the 41st Annual Meeting of European Society for Paediatric Endocrinology, Madrid, Spain, 2002, p 119
  38. Erlander MG, Tillakaratne NJK, Feldblum S, Patle N, Tobin AJ 1991 Two genes encode distinct glutamate decarboxylases. Neuron 7:91–100[CrossRef]<