ontogeny of the gnrh-, glutaminase- and glutamate decarboxylase-gene expression in the hypothalamus...

Ž .Developmental Brain Research 110 1998 105–114

Research report

Ontogeny of the GNRH-, glutaminase- and glutamatedecarboxylase-gene expression in the hypothalamus of female rats

Christian Roth a,), Sabine Leonhardt b, Kirsten Theiling b, Max Lakomek a, Hubertus Jarry b,Wolfgang Wuttke b

a Children’s Hospital, UniÕersity of Gottingen, Gottingen, Germany¨ ¨b DiÕision of Clinical and Experimental Endocrinology, Department of Obstetrics and Gynecology, UniÕersity of Gottingen, Gottingen, Germany¨ ¨

Accepted 23 June 1998

Abstract

Ž . Ž .Amino acid neurotransmitters like g-aminobutyric acid GABA and glutamate GLU are involved in the regulation of hypothalamicŽ .gonadotropin releasing hormone GnRH release. We investigated, whether there are changes of gene expression in the rat hypothalamus

for GnRH, GnRH receptor, as well as glutaminase and glutamate decarboxylase, two enzymes regulating neurotransmitter concentrationsŽ .of GLU and GABA in the brain during the ontogeny. After reverse transcription–polymerase chain reaction RT–PCR we used an

Ž .ELISA method to quantify PCR products. In 15-day old animals high plasma luteinizing hormone LH levels with pronounced variationswere found. In 25-day old animals LH values were low, whereas in 35-day old rats LH levels increased significantly indicating thereactivation of the GnRH-pulse generator at the beginning of puberty. In parallel to these changes, the mRNA levels of the GnRH

Žreceptor in the mediobasal hypothalamus were high at day 15, significantly lower at day 25 and again high at day 35 after birth ELISA.O.D. GnRH-R day 15: 0.46"0.07, day 25: 0.16"0.04, day 35: 0.36"0.04; p-0.01 , but no changes of GnRH receptor gene

expression were found in the preoptic area. The mRNA of GnRH in the preoptic area as well as mRNA levels of glutaminase andglutamate decarboxylase in the mediobasal hypothalamus and the preoptic area did not change during ontogeny. We conclude thathypothalamic GnRH receptors are involved in the characteristic changes of LH secretion patterns during sexual maturation. Majorchanges of GnRH receptor gene expression occurred in the mediobasal hypothalamus and correlated well with plasma LH levels, whereashypothalamic mRNA levels of GnRH, glutaminase and glutamate decarboxylase did not change within the different age groups. Thus theactivity of the GABA- and glutamatergic system during ontogeny may be regulated at the receptor or postreceptor level. q 1998 ElsevierScience B.V. All rights reserved.

Keywords: Puberty; Rat hypothalamus; mRNA; PCR ELISA; GnRH-pulse generator

1. Introduction

Previous studies demonstrate that amino acid neuro-transmitters such as the major excitatory amino acid neuro-

Ž .transmitter L-glutamate GLU and the major inhibitoryŽ .amino acid neurotransmitter g-aminobutyric acid GABA

are involved in the regulation of the hypothalamic go-Ž .nadotropin releasing hormone GnRH -pulse generator

) Corresponding author. Universitatskinderklinik, Robert-Koch-Str. 40,¨37075 Gottingen, Germany. Fax: q49-551-396231¨

w x7,10,17,19 . GABAergic and glutaminergic neurons pro-ject to GnRH-neurons. Furthermore neurointeractions ofGABAergic and glutaminergic neurons could be shownw x39 . The increased LH-secretion seen at the onset ofpuberty is dependent on a pulsatile release of GnRH from

w xGnRH neurons 7,18,28 . In the rat, cell bodies of GnRHŽ .neurons are located in the preoptic area POA of the

hypothalamus while their axons are mainly projecting intothe median eminence of the medial hypothalamus.

During sexual development changes of hypothalamicneurotransmitter concentrations were observed in rats in

w xvivo and in hypothalamus explants in vitro 1,14,17 . GLU,the most abundant excitatory neurotransmitter and N-

Ž .methyl-D-aspartate NMDA , an exogenous agonist of this

0165-3806r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII: S0165-3806 98 00102-3

( )C. Roth et al.rDeÕelopmental Brain Research 110 1998 105–114106

neurotransmitter, are able to stimulate GnRH release fromhypothalamic explants in prepubertal as well as in adult

w xmale and female rats in vitro 6,10,12,14 . During sexualmaturation of female rats NMDA receptor gene expression

w xon GnRH neurons increased 16 . Blockage of NMDAw xreceptors delayed the timing of puberty 41 . An increased

sensitivity to the stimulatory action of GLU on GnRHsecretion via NMDA receptors was shown at the onset of

w xpuberty 9 . In contrast, the hypothalamic GABAergiccontrol of GnRH secretion changes from a stimulatoryeffect in prepubertal to an inhibitory in peripubertal and

w xadult male and female rats 14,28,33,34 . In adult femalerats it could be demonstrated that muscimol, a GABA-Areceptor agonist, reduces serum LH and hypothalamicGnRH receptor mRNA levels. Therefore, the regulation ofGnRH receptor transcripts in the hypothalamus of ratsdepends indirectly on GABAergic neurotransmission

w xthrough GABA-A receptors 22,35 .Glutamine, released by astrocytes and taken up by

neurons, serves as a precursor for glutamate and GABAŽ .synthesis. Glutaminase GLS is the key enzyme that

catalyses the hydrolysis of glutamine to GLU and ammo-w xnia in neurons 2 . Astroglial cells can reuptake GLU from

the synaptic cleft, transforming GLU into glutamine byw xglutamine synthetase 8 . GLU is converted in the brain to

GABA by the rate-limiting enzyme glutamate decarboxyl-Ž . w xase GAD 43 . Two different GAD enzymes are present

Ž .in the brain of the rat: the apo-GAD primarily GAD65w xand the predominantly active GAD 24,40 . These two67

proteins are encoded by two different genes. GAD sup-65

posedly is specialized in synthesizing GABA for synapticrelease and might be able to respond quickly to suddenincreases in the demand for GABA, whereas GAD is67

probably regulated in a more long-term fashion. It hasbeen concluded that GAD is responsible for the synthesis67

of the GABA supporting tonic synaptic release of GABAw x25,26 . Furthermore, it has been shown that GAD , but67

w xnot GAD is regulated by sex steroids like estradiol 40 .65

For that reason we studied GAD mRNA levels. In rats67Ž .synaptic inputs of glutamate decarboxylase GAD -posi-

tive neurons were found on GnRH positive perikaryaw xlocated in the preoptic area 23 .

Very little data exists on the ontogenic changes of geneexpression of enzymes which are possibly involved in the

w xregulation of puberty 8 . Therefore the aim of the presentstudy was to investigate, whether there are ontogenicchanges of the mRNA’s from the GLU-synthesizing en-zyme GLS or the GABA-synthesizing enzyme GAD incomparison with hypothalamic GnRH and GnRH receptormRNAs. We studied rats with neonatally active GnRH-

Ž . Ž .pulse generator day 15 , prepubertal inactive day 25 andreactivated GnRH-pulse generator during early puberty atday 35 of life. This pattern is similar to that in humans,where pulsatile GnRH secretion in early infancy is fol-lowed by low GnRH secretion during childhood untilreactivation of pulsatile GnRH secretion occurs in puberty.

2. Materials and methods

2.1. Animals

Female Sprague–Dawley rats were maintained underŽstandardized conditions light on from 0700 h to 1900 h,.258C room temperature . They had free access to water

and were fed ad libitum. Litter size was adjusted to 12–14animals. Three groups of 15 animals were formed. Ratswere decapitated at age 15, 25 or 35 days, respectively.Furthermore ovariectomized adult rats were treated withGABA agonist muscimol to verify the validity of theA

Ž .PCR ELISA see Sections 2.4 and 2.7 .

2.2. Hormone parameters

Trunk blood was collected for determination of luteiniz-Ž .ing hormone LH and prolactin with established RIA-

Table 1Sequences of PCR primers and ELISA hybridization probes

XŽ .Probe Upstream position 5 -end Length Sequence Additional restriction sideX XŽ .GnRH 894 Exon I 20 5 -CACTATGGTCACCAGCGGGG-3 –fX XŽ .GnRH 1246 Exon III 24 5 -AGAGCTCCTCGCAGATCCCTAAGA-3 –rX XŽ . Ž .GnRH 1056 Exon II 28 5 - Bi -GGCCGCTGTTGTTCTGTTGACTGTGTGT-3 –bX XGnRH-Rec 53 32 5 -CAGGTAAGCTTGAAGCCCGTCCTTGGAGAAAT-3 HindIIIfX XGnRH-Rec 467 33 5 -ACTGAATTCGCGATCCAGGCTAATCGCCGCCAT-3 EcoRIrX XŽ .GnRH-Rec 340 24 5 - Bi -TGAGACTCTAATCGTCATGCCGCT-3bX XGAD 361 29 5 -GGCGGATCCAGTCGTCTTGTGAGCGCCTT-3 BamH1fX XGAD 681 29 5 -GGCAAGCTTCTCTACAGTCAACCAGGATC-3 HindIIIrX XŽ .GAD 531 26 5 - Bi -ACTGGAGGTGGTTGACATACTCCTCA-3bX XGLS 180 26 5 -GGCGGATCCAGGCACAGACATGGTTG-3 BamH1fX XGLS 651 26 5 -GGCAAGCTTCATCTCCAGTGTACGCA-3 HindIIIrX XŽ .GLS 344 29 5 - Bi -GCTTAATGCACTCCTGTGGCATGTATGAC-3b

Ž . Ž . Ž .Sequences of forward f and reverse r PCR primers and biotin-labelled b PCR ELISA hybridisation probes. The upstream positions are indicatedw x w x w x w xaccording to published cDNA sequences of Bond et al. 3 , Kaiser et al. 20 , Wyborsky et al. 43 and Banner et al. 2 .

Cursive characters indicate sequences of additional restriction sites.

( )C. Roth et al.rDeÕelopmental Brain Research 110 1998 105–114 107

Fig. 1. ELISA detection of cDNA’s of hypothalamic GnRH-R, GAD andŽ .GLS MBH after PCR. The cDNA samples were diluted down to 1:20

with distilled water before starting the PCR. With the ELISA method thecDNA’s could be detected over a wide concentration range.

methods. Serum LH levels were measured as describedw xpreviously 18 using an antiserum provided by Dr. Leong

Ž .University of Virginia, Charlottesville, VA, USA . Thelower detection limits were for LH 0.5 ngrml and forprolactin 0.7 ngrml, respectively.

2.3. Microdissection of brain areas

Brains were removed and rapidly frozen on dry ice. Thefrozen brains were mounted on a cryostat and coronallysectioned in 0.6 mm thick slices. The POA was punchedout from the brain slice using a Ø 1.0 mm hypodermicneedle according to the method of Palkovits and Brown-

w xstein 30 and coordinates corresponded to a: 6.3, l: 0.6, h:w x2.6 of the stereotaxic atlas of de Groot 11 . The MBH was

Žtotally dissected anterior border: optic chiasmi; posterior.border: mamilliary bodies, lateral 1 mm, depth 1 mm .

Tissue in comparable size was also dissected from tempo-ral cortex and corpora amygdaloidea according to de Grootw x11 .

2.4. IntracerebroÕentricular injection of muscimol

Adult female rats, mean weight 250 g, were ovariec-tomized under light ether anaesthesia 2 weeks prior toexperimentation. One week after ovariectomy, under

Žpentobarbital anaesthesia, a guide cannula o.d. 0.7 mm,.stainless steel tube was stereotaxically implanted into theŽ .left lateral ventricle A: 5.8, L: 1.5, V: 3.0 according to

w xthe atlas of de Groot 11 and fixed to the skull withanchor screws and dental cement. The lumen of the can-nula was closed with a mandril. Seven days later 10 nmol

Ž .of muscimol, dissolved in 2 ml 0.9% saline NaCl weremicroinjected through the guide cannula with the help of aHamilton syringe. The control rats received 2 ml of NaCl.

Ž .Two hours after intracerebroventricular i.c.v. injectionrats were decapitated and brains rapidly removed andfrozen on dry ice.

2.5. RNA-preparation

Micropunches of POA, tissues of MBH, cortex andŽ .corpora amygdaloidea were sonicated in 200 ml POA or

Ž .350 ml MBH, cortex, corpora amygdaloidea lysis buffer.Total RNA was prepared by using a RNA minipreparation

Ž .kit Rneasy Total RNA Kit, Quiagen, Hilden, Germany .In each age group the RNA was collected from 13 to 15animals. Extracted total RNA of MBH, cortex and corporaamygdaloidea was measured with an UV spectrophotom-eter at A 260 nm and each sample was diluted with H 0 to2

achieve a concentration of 100 ng RNArml before startingŽ . Ž .the reverse transcription RT with 1 ml 0.5 mg of oligo

Ž .dT-primer Gibco, Eggenstein, Germany and 8 ml H O,2. Ž .10 min, 708C . The following reaction mixture 20 ml

Ž . Žcontaining 1 ml 200 U reverse transcriptase Gibco,.Eggenstein, Germany , RNase inhibitor, 10 mM DTT, 0.5

ŽmM dNTP and RT buffer 50 mM Tris–HCl, pH 8.3, 75. ŽmM KCl, 3 mM MgCl was subjected to RT 10 min at2

.228C, 50 min at 428C, 10 min at 958C .

( )2.6. Polymerase chain reaction PCR

A total of 5 ml of the resulting cDNA was amplified byŽ2.5 U of Taq-DNA-polymerase Gibco, Eggenstein, Ger-

.many in 100 ml total volume containing 50 ml 2=ŽPCR-buffer 20 mM Tris–HCl, pH 8.3, 100 mM KCl, 3

.mM MgCl , 0.5 mM dNTP mixture, 1% gelatin by 502Ž .pmol of the following PCR primers Table 1 . PCR proce-

dure was performed in a Biometra automated thermocyclerŽ .1 min 948C, 1 min 608C, 2 min 728C over 28 to 38cycles. For quantitative determination of mRNA levels itwas necessary to avoid saturation in the PCR. Therefore,different numbers of PCR cycles were tested in precedingexperiments. To prevent or destroy secondary structures,the PCR was continued with seven cycles of 1 min at788C, 1 min at 608C and 2 min at 728C followed by a final

Fig. 2. The ELISA detection showed a small variance between threeŽ .different dilution rows A, B, C . After the PCR the amplified DNA was

diluted down to 1:40 with distilled water before starting the ELISA. Inthis dilution experiment a pituitary cDNA sample was used after GnRHreceptor PCR amplification.


Fig. 3. Values of triplicate samples in three different dilutions of hypotha-Ž .lamic cDNA. The sample was diluted 1:1, 1:2 or 1:5 100%, 50%, 20%

Ž .before the glutaminase GLS -PCR was started. There was a clearcorrelation between the values obtained by the ELISA method compared

Ž .with the conventional densitometric gel analysis r s0.96 .

Ž .step at 798 for 3 min only for GnRH-R . The samples ofthe different age groups were investigated in the same RTand PCR procedure to allow the comparison of mRNAlevels between the age groups. The PCR products had a

Ž . Ž .length of 441 bp GnRH receptor , 460 bp GnRH , 487 bpŽ . Ž .GLS and 340 bp GAD .

2.7. EÕaluation by using a PCR ELISA technique

The PCR products were labelled with digoxigenin dur-Žing the amplification process PCR ELISA, Dig Labeling,

.Boehringer, Mannheim, Germany . A specific captureprobe, which is complementary to the inner part of theamplification product was labelled at the 5X-end with biotinto allow immobilization of the hybrid on a streptavidincoated microtiter plate surface. These bound hybrids weredetected by an anti-digoxigenin peroxidase conjugate andfollowed by colorimetric measurement of the peroxidase

Ž X Žsubstrate ABTS 2,2 -azino-di- 3-ethylbenzthiazoline sul-.fonic acid which allowed a semiquantitative evaluation

and comparison between the different age groups. The

Žextinction was measured using a photometer measure.filter 405 nm, reference filter 492 nm . Sequences of the

5X-biotin labelled capture probes are given in Table 1.Samples of each PCR ELISA were evaluated by con-

ventional gel electrophoresis, ethidium bromide stainingand densitometric analysis of the gel to verify that thePCR-products we looked for were present and only onesingle band of the predicted size was detectable. Forsemiquantitative detection of mRNA levels both conven-tional RT–PCR methods with densitometric evaluation andRT–PCR ELISAs were used. The method of densitometricevaluation is simple and sensitively showed changes inmRNA levels, but was of limited use in large series ofinvestigated samples. Therefore, we used a PCR ELISAmethod in which digoxigenin labelled PCR products wereanalyzed by hybridization to a specific capture probe. For

Ž .each PCR ELISA GnRH, GnRH receptor, GLS, GAD wedeveloped the specific hybridization capture probe, whichis complementary to the inner part of the amplification

Ž .product Table 1 . The PCR ELISAs allowed us to mea-Ž .sure a wide range of mRNA levels O.D. 0.1–1.5 and

showed a small intra-assay variance. Statistical evaluationof samples, which were analyzed by both methods, gelelectrophoresis and PCR ELISA, showed a clear correla-tion between the two methods.

To demonstrate the validity of the PCR ELISA to detectchanges of hypothalamic GnRH mRNA levels, ovariec-tomized adult rats were treated with GABA agonistA

muscimol. Two hours after intracerebroventricular injec-tion of 10 nmol muscimol, GnRH nRNA levels in thepreoptic were measured. To quantify the generated PCRproducts gel electrophoresisrdensitometry and ELISA de-tection were used.

2.8. Statistical eÕaluation and mathematical calculations

Unless otherwise stated, results are presented as mean"S.E.M. The differences arising between the groups weretested for significance using one way analysis of variancefollowed by a two tailed t-test for unpaired samples

Ž . Ž .Fig. 4. Gel of the glutaminase GLS -PCR at day 15, 25 and 35 in triplicates. The cDNA sample of day 35 was diluted 1:1, 1:2 or 1:5 100%, 50%, 20%Ž .before the PCR was started see Fig. 3 . The undiluted blots differ not significantly between the age groups. The blots of the diluted samples of day 35

were markedly weaker.


Fig. 5. The blood LH levels of rats showed a high variation at day 15,lower levels at day 25 and later, on day 35, there was a significant

Ž .increase Mann–Whitney .

Ž .Bonferroni’s multiple t-test . For nonparametric parame-ters we used a two-tailed Mann–Whitney test as indicated.

w xA p-0.05 was considered to be significant 4 .

3. Results

3.1. RT–PCR ELISA method

For semiquantitative detection of mRNA levels bothconventional RT–PCR methods with densitometric evalua-tion and RT–PCR ELISAs were used. The method ofdensitometric evaluation is simple and sensitively showedchanges in mRNA levels, but was of limited use in largeseries of investigated samples. For that reason we used a

PCR ELISA method in which digoxigenin labelled PCRproducts were analyzed by hybridization to a specific

Žcapture probe. For each PCR ELISA GnRH, GnRH recep-.tor, GLS, GAD we developed the specific hybridization

capture probe, which is complementary to the inner part ofthe amplification product. The PCR ELISA allowed us to

Žmeasure a wide range of mRNA levels O.D. 0.1–1.5, see. Ž .Fig. 1 and showed a small intra-assay variance Fig. 2 .

Statistical evaluation of samples, which were analyzed byboth methods, gel electrophoresis and PCR ELISA, showed

Ža clear correlation between the two methods Figs. 3 and.4 .

3.2. Hormone parameters

During ontogeny LH showed a high variation in rats atday 15 indicating the known pulsatile secretion at this age.We found low levels of LH on day 25 in juvenile rats.During early puberty at day 35 there was a significantincrease of plasma LH due to the beginning of reactivationof pulsatile GnRH release from hypothalami. Characteristi-cally prolactin showed a significant increase from day 15

Ž .to day 35 Figs. 5 and 6 .

3.3. ComparatiÕe study of hypothalamic gene expression ofGnRH, GnRH receptor, GLS and GAD

There were no significant changes of mRNA levels ofGnRH, GnRH receptor, GLS and GAD in the preoptic areaand of GLS and GAD in the mediobasal hypothalamusbetween the three age groups. In contrast the GnRH recep-tor mRNA levels showed a significant decrease at day 25

Fig. 6. Serum prolactin values increased significantly during ontogeny from day 15 to day 25 and from day 25 to day 35. Mean valuesqS.E.M. are given,Ž .p-0.0001 Mann–Whitney .


Ž . Ž . Ž . Ž .Fig. 7. The mRNA levels of GnRH, GnRH receptor GnRH-R , glutaminase GLS and glutamate decarboxylase GAD in the preoptic area POA of ratsŽ .at days 15, 25 and 35 are shown. No significant differences were found between the age groups. Mean optical density O.D. valuesqS.E.M. are given.

Ž .in the MBH p-0.01 . GnRH gene expression was de-Žtectable only in the POA and not in the MBH Figs. 7 and

.8 .We also studied GnRH receptor, GLS and GAD mRNA

levels in the cortex and the corpora amygdaloidea, whichare thought not to be involved in the onset of puberty. Inthese brain areas the gene expression of GnRH receptor,

ŽGLS and GAD did not change during ontogeny data not.shown .

3.4. GnRH mRNA leÕels in the preoptic area after intrac-erebroÕentricular injection of muscimol

In ovariectomized adult rats intracerebroventricular in-jection of 10 nmol muscimol caused a significant reduction

Ž . Ž .Fig. 8. In the mediobasal hypothalamus MBH levels of GnRH-receptor GnRH-R mRNA on day 25 were significantly decreased compared with days 15Ž . Ž .and 35 p-0.01 . The mRNA levels of GLS and GAD were unchanged in the different age groups. Mean optical density O.D. valuesqS.E.M. are

given.


Ž . ŽFig. 9. GnRH levels in the POA left part and mean LH-secretion right.part of the graph of ovariectomized rats. Animals were injected with 10

nmol muscimolr2 ml 0.9% NaCl or saline into the lateral ventricle andkilled 2 h later by decapitation. Note the significant inhibition of LHrelease induced by muscimol. With both, gel electrophoresisrdensitom-etry and ELISA detection; a significant reduction of GnRH mRNA levelsin the POA was observed. Mean valuesqS.E.M., ns10 rats per group,) p-0.05 vs. saline injected rats.

of GnRH mRNA levels in the preoptic area. This reductionwas detectable with both techniques to similar extent. LHsecretion was clearly reduced as a consequence of the

Ž .muscimol induced inhibition of GnRH release Fig. 9 .

4. Discussion

We investigated the gene expression of GnRH, GnRHreceptor and two enzymes, GAD and GLS, which areinvolved in the regulation of the two major amino acidneurotransmitters, GABA and GLU by using two methods,conventional RT–PCR with densitometric evaluation aftergel electrophoresis and RT–PCR ELISAs. During on-togeny we found characteristic changes of serum LH levelsand GnRH receptor gene expression in the MBH. Hypotha-lamic mRNA levels of GnRH, GAD and GLS did notchange within the different age groups.

We used a PCR ELISA method which showed clearadvantages in the analysis of large series of samplescompared with the standard densitometric evaluation of agel. For each PCR ELISA we developed the specifichybridization capture probe. The PCR ELISA allowed usto measure a wide range of mRNA levels and showed asmall intra-assay variance.

The high serum LH levels at the age of 15 days indicatean active GnRH pulse generator, which becomes arrestedlater on, as seen by low LH levels on day 25. High LHvalues on day 35 indicate the reactivation of the GnRHpulse generator at the start of puberty. These changes

w xconfirm our previous results 32,42 . In this study wemeasured LH levels only in the trunk blood, i.e., at onlyone time point, but from previous investigations we knowthat pulsatile LH secretion is present in neonatal rats untilthe age of 20 days. In juvenile rats from 20–35 days of agethere are low LH levels and no pulsatile LH-secretion,whereas pulsatile LH-secretion with high LH peaks starts

w xto occur in early puberty at day 35 42 .Serum prolactin levels increased during ontogeny char-

w xacteristically confirming previous results 42 . In femalerats it could be demonstrated that the influence of GABAon GnRH and prolactin release changes from a stimulatoryin prepubertal to an inhibitory effect in peripubertal rats,indicating that complex neuroendocrine mechanisms of theGABAergic system influencing the prolactin secretion areinvolved in GnRH secretion and the onset of pubertyw x14,26,29 .

Although there were characteristic changes of LH secre-tion from neonatal to juvenile and from juvenile to puber-tal rats, we could not find significant changes of GnRHgene expression in the POA during ontogeny. This issurprising since perikarya of GnRH neurons in the rat arelocated in this area and changes in GnRH content and geneexpression have been described in adult rats depending on

w xthe stage of estrous cycle 15 . To demonstrate that theseunchanged levels are not due to a methodical problem, weinjected ovariectomized rats with the GABA agonistA

muscimol into the lateral ventricle. LH secretion and GnRHmRNA levels were measured 2 h after injection. As re-

w xported previously 22 , muscimol caused a pronounceddecrease of LH secretion which was accompanied by asignificant reduction of GnRH mRNA levels in the POA.With both methods used to quantify mRNA levels, i.e., gelelectrophoresisrdensitometry or ELISA, the reduction wasobserved to similar extent. Thus, the ELISA is a suitabletechnique which is sensitive enough to observe changes inGnRH mRNA levels in individual animals and thereforethe missing changes of GnRH mRNA levels during theprepubertal life span are not due a technical probleminherent to the detection method.

In the MBH we measured high mRNA levels of GnRHŽ .receptor in neonatal animals day 15 , low levels in prepu-

Ž . Žbertal day 25 and again high levels in pubertal rats day.35 , showing nearly the same pattern as the LH levels. The

change of GnRH receptor gene expression in the MBH ispossibly involved in the peripubertal LH increase. How-ever, the decrease in GnRH receptor mRNA in juvenilerats and the subsequent increase after the onset of pubertycould be looked as a consequential rather than a causalevent. Superfusion of rat pituitary cells with pulses ofGnRH in vitro results in an increase of GnRH receptor

w xmRNA levels 21 . In addition to gonadotroph cells, GnRHreceptors have been characterized in gonads, placenta andin various brain tissues as well as in immortalized hypotha-

w xlamic neurons 38 . The localization of GnRH receptorsand GnRH receptor mRNA in the hypothalamus suggests


that the native GnRH decapeptide could be involved in anw xautofeedback effect 5 .

Previous studies showed that the GnRH release fromexplanted rat MBH fragments were influenced by neuro-transmitters such as GLU, NMDA, GABA and taurinw x6,8,13,14 . These fragments contain GnRH axon termi-nals, but no GnRH perikarya which are located in thePOA. It was interesting to note that during ontogenyGnRH receptor gene expression in the MBH was de-creased while it was unchanged in the POA. It is knownthat in the pituitary missing exposure of GnRH receptor toits ligand causes downregulation of GnRH receptors. Un-der the assumption that hypothalamic GnRH receptors areregulated in a similar way, the phenomenon of a differingexpression of the GnRH receptors in the MBH might be

w xexplained as follows. Bourguignon et al. 7 have sug-gested that various neurotransmitters modulate mediobasalGnRH release via presynaptic interactions. If, via theseinteractions, GnRH release in the MBH is reduced thenGnRH receptor gene expression in the MBH should bedownregulated. However, preoptic GnRH release remainspulsatile and therefore, no downregulation of the GnRHreceptor occurs in the POA. Thus, GnRH receptors in theMBH appear to be involved in the regulation of GnRHrelease and major regulation of GnRH secretion resides in

w xthe MBH, confirming results from Bourguignon et al. 7 ,whereas GnRH receptors in the POA appear to be involvedin the pulse generation of GnRH as it could be shown by

w xSeong et al. 35 .Puberty can be caused by either reactivation of the

excitatory system or desinhibition of the inhibitory system.Administration of either the GLU agonist NMDA or theGABA-A antagonist bicucullin in prepubertal monkeys

w xand rats advanced the onset of puberty 12,26,31 .We speculated that GABA may be reduced before onset

of puberty and that this decrease is accompanied by devel-opmental changes in the gene expression of the biosyn-thetic enzyme GAD. GABA plays a role in the tonicinhibition of GnRH release in prepubertal rats mediated by

w xGABA-A receptors 28,36 . In prepubertal monkeys it hasalso been shown that GnRH neurons are tonically inhibitedby endogenous GABA, and higher GABA release in thestalk median eminence were found in prepubertal com-

w xpared with pubertal monkeys 26 . Hypothalamic GABA-Aw xreceptor levels are known to be steroid regulated 38 .

Blocking of GABA-A receptors by bicuculline in prepu-bertal animals results in elevated LH levels and might lead

w xto precocious puberty 26,27 . It seems that GABA is themost important neurotransmitter in inducing pulsatile hy-pothalamic GnRH secretion which leads to puberty andthat GABAergic neurons are the key component in thehypothalamic network comprising the GnRH pulse genera-tor. Hypothalamic GAD mRNA levels, however, did notchange in female rats between the age of 15 to 35 days.Possibly, changes occur at earlier or later age. Investiga-tions of the gene expression of the GAD family in the

cervical region of rat spinal cords revealed an increase ofmRNA levels from embryonic to neonatal rats but no

w xchanges of gene expression thereafter 37 .GLS, the enzyme controlling GLU biosynthesis from

glutamine, was found to be enriched in brain neuronsw xwhich release glutamate as a neurotransmitter 2 . It has

been shown that GLU is strikingly involved in the regula-tion and the characteristic changes of GnRH release duringontogeny. In a previous study it was suggested that inretrochiasmatic hypothalamic explants of male rats GLSshows an increased activity after the onset of puberty,when the frequency of pulsatile GnRH secretion is in-

w xcreased as well 8 . Although there are evident changes ofhypothalamic neurotransmitter contents during ontogeny,we could not find changes of hypothalamic GLS and GADgene expression. This might be due that we looked at onestep of the biosynthesis of GABA or glutamate, but it isnot out of the question that there are changes in the releaserates of these neurotransmitters.

Our interest was focused on mRNA levels of hypothala-mic nuclei, which may be predominantly involved in theregulation of GnRH secretion. Additionally, we looked tothe gene expression of the GnRH receptor, GLS and GADother brain areas such as cortex and corpora amygdaloidea,which are thought not to be involved in the onset ofpuberty. The finding that in these brain areas the geneexpression of GnRH receptor, GLS and GAD are notchanging during ontogeny underline the specificity of lowGnRH receptor gene expression in the MBH in juvenilerats and show that this is not a generalized pattern includ-ing other brain areas.

In conclusion, corresponding with the well known pul-satile LH-secretion in neonatal and pubertal rats, mRNAlevels of GnRH receptor in MBH are higher in neonataland pubertal female rats compared with those of juvenilerats. Although there was a high gonadotropin secretion inneonatal and pubertal compared to juvenile rats, which wasrelated to high hypothalamic GnRH secretion, there wereno changes in mRNA levels of GnRH in the POA. ThemRNA levels of GAD and GLS in POA and MBH did notchange from neonatal, juvenile nor pubertal rats as well.

Acknowledgements

This study was supported by the Deutsche Forschungs-Ž .gemeinschaft Grant Ja 398r4-2 . RIA material was kindly

provided by the National Hormone and Pituitary ProgramŽ .Dr. A.F. Parlow .

References

w x1 P. Arias, H. Jarry, S. Leonhardt, J.A. Moguilevsky, W. Wuttke,Estradiol modulates the LH release response to N-methyl-D-aspar-tate in adult female rats: studies on hypothalamic luteinizing hor-


mone-releasing hormone and neurotransmitter release, Neuroen-Ž .docrinology 57 1993 710–715.

w x2 C. Banner, J.J. Hwang, R.A. Shapiro, R.J. Wenthold, Y. Nakatani,K.A. Lampel, J.W. Thomas, D. Huie, N.P. Curthoys, Isolation of a

Ž .cDNA for rat brain glutaminase, Molecular Brain Res. 3 1988247–254.

w x3 C.T. Bond, J.S. Hayflick, J.H. Seeburg, J.P. Adelman, The ratgonadotropin-releasing hormone: SH locus: structure and hypothala-

Ž .mic expression, Mol. Endocrinol. 3 1989 1257–1262.w x4 Bonferroni Multiple T-Test. Instat Programm Graph PAD, San

Diego, CA, USA.w x5 J.P. Bourguignon, A. Gerard, G. Debougnoux, J. Rose, P. Franchi-´

Ž .mont, Pulsatile release of gonadotropin-releasing hormone GnRHfrom the rat hypothalamus in vitro and glucose dependency and

Ž .inhibition by superactive GnRH analogs, Endocrinology 121 1987993–999.

w x6 J.P. Bourguignon, A. Gerard, J. Mathieu, J. Simons, P. Franchimont,´Pulsatile release of gonadotropin-releasing hormone from hypothala-mic explants is restrained by blockade of N-methyl-D-aspartate

Ž .receptors, Endocrinology 125 1989 1090–1096.w x7 J.P. Bourguignon, A. Gerard, M.L. Alvarez Gonzalez, P. Franchi-´

mont, Neuroendocrine mechanism of onset of puberty, J. Clin.Ž .Invest. 90 1992 1736–1744.

w x8 J.P. Bourguignon, A. Gerard, M.L. Alvarez Gonzalez, G. Purnelle,´P. Franchimont, Endogenous glutamate involvement in pulsatilesecretion of gonadotropin-releasing hormone: evidence from effect

Ž .of glutamine and developmental changes, Endocrinology 136 1995911–915.

w x9 J.P. Bourguignon, A. Gerard, M.L. Alvarez Gonzalez, P. Franchi-´mont, Role of excitatory amino acids in the physiological neuroen-

Ž .docrine mechanism of puberty, Frontiers in Endocrinol. 10 1995215–223.

w x10 D.W. Brann, V.B. Mahesh, Excitatory amino acids: function andsignificance in reproduction and neuroendocrine regulation, Fron-

Ž .tiers in Neuroendocrinol. 15 1994 3–49.w x11 J. de Groot, The rat forebrain in stereotaxic coordinates, Verh. K.

Ž .Akad. Naturk. 2 1959 1–40.w x12 A. Donoso, F. Lopez, A. Negro-Vilar, Glutamate receptors of the

non-N-methyl-D-aspartatic acid type mediate the increase in luteiniz-ing hormone-releasing hormone release by excitatory amino acids,

Ž .Endocrinology 126 1990 414–420.w x13 A.O. Donoso, F.J. Lopez, A. Negro-Vilar, Cross-talk between exci-

tatory and inhibitory amino acids in the regulation of luteinizingŽ .hormone-releasing hormone secretion, Endocrinology 3 1992

1559–1561.w x14 C. Feleder, H. Jarry, S. Leonhardt, W. Wuttke, J.A. Moguilevsky,

The GABAergic control of gonadotropin-releasing hormone secre-tion in male rats during sexual maturation involves effects onhypothalamic excitatory and inhibitory amino acid systems, Neu-

Ž .roendocrinology 64 1996 305–312.w x15 A.C. Gore, J.L. Roberts, Regulation of gonadotropin-releasing hor-

mone gene expression in the rat during the luteinizing hormoneŽ .surge, Endocrinology 136 1995 889–896.

w x16 A.C. Gore, T.J. Wu, J.J. Rosenberg, J.L. Roberts, Gonadotropinreleasing hormone and NMDA receptor gene expression and colo-calization change during puberty in female rats, J. Neurosci. 16Ž .1996 5281–5289.

w x17 D. Goroll, P. Arias, W. Wuttke, Ontogenic changes in the hypothala-mic levels of amino acid neurotransmitters in the female rat, Dev.

Ž .Brain Res. 77 1994 183–188.w x18 H. Jarry, S. Leonhardt, W. Wuttke, g-Aminobutyric acid neurons in

the preopticranterior hypothalamic area synchronize the phasic ac-tivity of the gonadotropin-releasing hormone pulse generator in

Ž .ovariectomized rats, Neuroendocrinology 53 1991 261–267.w x19 H. Jarry, S. Leonhardt, T. Schwarze, W. Wuttke, Preoptic rather

than mediobasal hypothalamic amino acid neurotransmitter release

regulates GnRH secretion during the estrogen-induced LH surge inŽ .the ovariectomized rat, Neuroendocrinology 62 1995 479–486.

w x20 U.B. Kaiser, D. Zhao, G.R. Cardona, W.W. Chin, Isolation andcharacterization of cDNAs encoding the rat pituitary gonadotropin-releasing hormone receptor, Biochem. Biophys. Res. Commun. 189Ž .1992 1645–1652.

w x21 U.B. Kaiser, A. Jakubowiak, A. Steinberger, W.W. Chin, Regulationof rat pituitary gonadotropin-releasing hormone receptor mRNA

Ž .levels in vivo and in vitro, Endocrinology 133 1993 931–934.w x22 S. Leonhardt, Y.J. Seong, K. Kim, Y. Thorun, W. Wuttke, H. Jarry,

Activation of central GABA-A but not of GABA-B receptors rapidlyreduces pituitary LH release and GnRH gene expression in thepreopticranterior hypothalamic area of ovariectomized rats, Neu-

Ž .roendocrinology 61 1995 655–662.w x23 C. Leranth, N.J. Mac Lusky, H. Sakamoto, M. Shanabrough, F.

Naftolin, Glutamic acid decarboxylase-containing axons synapse onLHRH neurons in the rat medial preoptic area, Neuroendocrinology

Ž .40 1985 536–539.w x24 D.L. Martin, S.B. Martin, S.J. Wu, N. Espina, Regulatory property

Ž .of brain glutamate decarboxylase GAD : the apoenzyme of GAD ispresent principally as the smaller of two molecular forms of GAD in

Ž .brain, Neuroscience 7 1991 91–100.w x25 M.M. McCarthy, D.B. Masters, K. Rimvall, S. Schwartz-Giblin,

D.W. Pfaff, Intracerebral administration of antisense oligodeoxynu-cleotides to GAD and GAD mRNAs modulate reproductive65 67

Ž .behavior in the female rat, Brain Res. 636 1994 209–220.w x26 D. Mitsushima, D.L. Hei, E. Terasawa, g-Aminobutyric acid is an

inhibitory neurotransmitter restricting the release of luteinizing hor-mone-releasing hormone before the onset of puberty, Proc. Natl.

Ž .Acad. Sci. USA 91 1994 395–399.w x27 D. Mitsushima, F. Kimura, The maturation of GABA receptor-A

mediated control of luteinizing hormone secretion in immature maleŽ .rats, Brain Res. 748 1997 258–262.

w x28 J.A. Moguilevsky, S. Carbone, B. Szwarcfarb, D. Rondina, Sexualmaturation modifies the GABAergic control of gonadotrophin secre-

Ž .tion in female rats, Brain Res. 563 1991 12–16.w x29 J.A. Moguilevsky, S. Carbone, B. Szwarcfarb, Changes in the effect

of g-aminobutyric acid on prolactin secretion during sexual matura-Ž .tion in female rats, Endocrinology 131 1992 458–462.

w x30 M. Palkovits, M.J. Brownstein, Microdissection of brain areas by theŽ .punch technique, in: A.C. Cuello Ed. , Brain Microdissection Tech-

niques, Vol. 2, Wiley, New York, 1983, 1–36.w x31 T.M. Plant, V.L. Gay, G.R. Marshall, M. Arslan, Puberty in mon-

keys is triggered by chemical stimulation of the hypothalamus, Proc.Ž .Natl. Acad. Sci. USA 86 1989 2506–2510.

w x32 C. Roth, H. Jung, K. Kim, P. Arias, J. Moguilevski, H. Jarry, S.Leonhardt, W. Wuttke, Involvement of gamma amino butyric acidŽ .GABA in the postnatal function of the GnRH pulse generator asdetermined on the basis of GnRH and GnRH-receptor gene expres-sion in the hypothalamus and the pituitary, Exp. Clin. Endocrinol.

Ž .Diabetes 105 1997 353–358.w x33 P. Scachi, S. Carbone, B. Szwarcfarb, D. Rondina, W. Wuttke, J.A.

Moguilevski, Interactions between GABAergic and serotoninergicsystems with excitatory amino acid neurotransmission in the hy-pothalamic control of gonadotropin secretion in prepubertal female

Ž .rats, Dev. Brain Res. 105 1998 51–58.w x34 B. Szwarcfarb, S. Carbone, M.L. Stein, J. Medina, J. Moguilevsky,

Sexual differences in the effect of the GABAergic system on LHsecretion and the hypothalamic ontogenesis of GABA-A receptors in

Ž .prepubertal rats, Brain Res. 646 1994 351–355.w x35 Y.J. Seong, H. Jarry, S. Kuhnemuth, S. Leonhardt, W. Wuttke, K.¨

Kim, Effect of GABAergic compounds on gonadotropin-releasinghormone receptor gene expression in the rat, Endocrinology 136Ž .1995 2587–2593.

w x36 S.T. Smith, C. Brennan, A.S. Clark, L.P. Henderson, GABA-Areceptor-mediated responses in the ventromedial nucleus of the


hypothalamus of female and male neonatal rats, NeuroendocrinologyŽ .64 1996 103–113.

w x37 R. Somogyi, X. Wen, W. Ma, J.L. Barker, Developmental kineticsof GAD family mRNAs parallel neurogenesis in the rat spinal cord,

Ž .J. Neurosci. 15 1995 2575–2591.w x38 S.S. Stojilkovic, J. Reinhardt, K.J. Catt, Gonadotropin-releasing

hormone receptors: structure and signal transduction pathways, En-Ž .docrine Rev. 15 1994 462–499.

w x39 K.K. Thind, P.C. Goldsmith, Glutamate and GABAergic neurointer-actions in the monkey hypothalamus: a quantitative immunomorpho-

Ž .logical study, Neuroendocrinology 61 1995 471–485.w x40 R. Unda, D.W. Brann, V.B. Mahesh, Progesterone suppression of

Ž .glutamic acid decarboxylase GAD mRNA levels in the preoptic67

area: correlation to the luteinizing hormone surge, Neuroendocrinol-Ž .ogy 62 1995 562–570.

w x Ž .41 H.F. Urbanski, S.R. Ojeda, A role for N-methyl-D-aspartate NMDAreceptors in the control of LH secretion and initiation of female

Ž .puberty, Endocrinology 126 1990 1774–1776.w x42 W. Wuttke, H. Jarry, S. Leonhardt, P. Arias, J.A. Moguilevsky, The

ontogeny of the LHRH pulse generator, Frontiers in Endocrinol. 10Ž .1995 1–12.

w x43 R.J. Wyborski, R.W. Bond, D.I. Gottlieb, Characterization of acDNA coding for rat glutamic acid decarboxylase, Mol. Brain Res. 8Ž .1990 193–198.

ontogeny of the gnrh-, glutaminase- and glutamate decarboxylase-gene expression in the hypothalamus...

Documents