Growth-Associated Protein-43 Messenger Ribonucleic Acid ...

0013-7227/00/$03.00/0 Endocrinology Copyright © 2000 by The Endocrine Society

Vol. 141, No. 5 Printed in U.S.A.

Growth-Associated Protein-43 Messenger Ribonucleic Acid Expression in Gonadotropin-Releasing Hormone Neurons during the Rat Estrous Cycle VINCENT PREVOT, SEBASTIEN BOURET, DOMINIQUE CROIX, GERARD ALONSO, LOTHAR JENNES, VALERIE MITCHELL, ARYEH ROUTTENBERG, AND JEAN-CLAUDE BEAUVILLAIN INSERM, U-422, Institut Fe´de´ratif de Recherches 22, Unite´ de Neuroendocrinologie et de Physiopathologie Neuronale (V.P., S.B., D.C., V.M., J.-C.B.), 59045 Lille, France; and INSERM U-336, University of Montpellier II (G.A.), 34095 Montpellier, France; University of Kentucky College of Medicine (L.J.), Lexington, Kentucky 40536; and Cresap Neuroscience Laboratory, Northwestern University (A.R.), Evanston, Illinois 60208 ABSTRACT We have shown previously at the ultrastructural level that morphological changes occur in the external zone of the median eminence allowing certain GnRH nerve terminals to contact the pericapillary space on the day of proestrus. The present study was designed to determine whether the intrinsic determinant of neuronal outgrowth, growth-associated protein-43 (GAP-43), was expressed in GnRH neurons of adult female rats, and whether its expression varied throughout the estrous cycle. To accomplish this, we perfusion-fixed groups of adult female rats at 0800 and 1600 h on diestrous day 2 (diestrous II), at 0800 h and 1600 h on proestrus, and at 0800 and 1600 h on estrus (n ⫽ 4 rats/group) and used double labeling in situ hybridization and quantification to compare the levels of GAP-43 messenger RNA (mRNA) in cells coexpressing GnRH mRNA. GnRH mRNA was detected with an antisense complementary RNA (cRNA) probe labeled with the hapten digoxigenin, whereas the GAP-43 cRNA probe was labeled with 35S and detected by autoradiography. In addition, GAP-43 protein was identified with immunohistochemistry in the median eminence. The results show that many GnRH neurons ex-

T

HE PREOVULATORY surge of LH is initiated by an abrupt increase in the release of GnRH (1, 2). This event is linked to an activation of the GnRH perikarya in the preoptic area (3, 4) and to morphological changes at the site of GnRH release, i.e. the median eminence, allowing GnRH axon terminals to contact the perivascular space directly (5, 6). Actual physical contacts between GnRH nerve endings and the parenchymatous basal lamina surrounding the perivascular space can be seen only during proestrus, whereas glial elements are interposed during the other stages of the cycle. These observations suggest that GnRH neurosecretory endings could be able to sprout spontaneously and thus reach the basal lamina to facilitate the secretion of the neurohormone into the portal blood during the preovulatory surge. As the regulation of neurite outgrowth and structural plasticity involves the expression of growth-associated factors, Received July 21, 1999. Address all correspondence and requests for reprints to: Dr. Vincent Prevot, INSERM, U-422, Unite´ de Neuroendocrinologie et de Physiopathologie Neuronale, 59045 Lille Cedex, France. E-mail: prevot@ lille.inserm.fr.

pressed GAP-43 mRNA and that GAP-43 protein was present in many GnRH axon terminals in the outer layer of the median eminence. The number of GnRH neurons expressing GAP-43 mRNA was significantly higher on proestrus (64 ⫾ 5%) than on diestrous II (40 ⫾ 2%; P ⬍ 0.001) or on estrus (45 ⫾ 8%; P ⬍ 0.05), and the GAP-43 mRNA levels in GnRH neurons also varied as a function of time of death during the estrous cycle. The GAP-43 mRNA levels in GnRH neurons were higher on proestrus and estrus than on diestrous II (P ⬍ 0.05). These data show that 1) GAP-43 is expressed in adult GnRH neurons; 2) GAP-43 mRNA expression in GnRH neurons fluctuates during the estrous cycle; and 3) GAP-43 mRNA content in GnRH neurons is highest on the day of proestrus, before and during the onset of the LH surge. These observations suggest that the increased GAP-43 mRNA expression in GnRH neurons on the day of proestrus could promote the outgrowth of GnRH axon terminals to establish direct neurovascular contacts in the external zone of the median eminence and thus facilitate GnRH release into the pituitary portal blood. (Endocrinology 141: 1648 –1657, 2000)

the present study was designed to determine whether growth associated protein-43 (GAP-43) was expressed in adult GnRH neurons and whether GAP-43 protein was present in GnRH axon terminals in the median eminence. GAP-43 stimulates neurons to sprout new terminals even in the adult and can therefore be considered an intrinsic determinant of the neuron’s growth state (7, 8). The present study examines the periovulatory patterns of GAP-43 gene expression in GnRH neurons to determine the temporal relationships with the GnRH/LH surge. Our results show that adult GnRH neurons express GAP-43 messenger RNA (mRNA), that GAP-43 mRNA expression in GnRH neurons varies during the rat estrous cycle, and that GAP-43 protein is transported to the GnRH axon terminals. Materials and Methods Animals Twenty-four adult female Wistar rats (235–250 g; CERJ, Saint Berthevius, France) were maintained on a 14-h light, 10-h dark cycle (lights on at 0500 h), with food and water available ad libitum. The estrous cycle was monitored by daily inspection of vaginal cytology. After at least 2 complete 4-day cycles, the 24 animals were divided into 6 groups,

1648

GAP-43 mRNA EXPRESSION IN GnRH NEURONS DURING RAT ESTROUS CYCLE

1649

representing different phases of the periovulatory period: diestrous day 2 (diestrous II) 0800 h (n ⫽ 4), diestrous II 1600 h (n ⫽ 4), proestrus 0800 h (n ⫽ 4), proestrus 1600 h (n ⫽ 4), estrus 0800 h (n ⫽ 4), and estrus 1600 h (n ⫽ 4). Groups were chosen to coincide with times when plastic changes in the GnRH nerve terminal level in the median eminence have been described (proestrus 0800 h and proestrus 1600 h) (5), and when the GnRH release in the portal blood reaches its peak (proestrus 1600 h) (9) as well as for comparison with other stages of the estrous cycle (diestrus 0800 and 1600 h; estrus 0800 and 1600 h) when GnRH release is maintained at a basal rate (9) and when no morphological change occurs at the GnRH nerve endings (5). The time points were also chosen to represent cycle stages in which the steroid environment is different (5, 10): before preovulatory estrogen secretion (diestrus 0800 and 1600 h), during the preovulatory estrogen surge (proestrus 0800 and 1600 h), during high progesterone (proestrus 1600 h) secretion, and after preovulatory estrogen and progesterone secretion (estrus 0800 and 1600 h).

Tissue preparation Animals were anesthetized with ketamine (20 mg/kg) and xylazine (0.2 ml/kg). Blood was collected, and plasma was stored at ⫺20 C for the estradiol, progesterone, and LH RIAs. Animals were perfused transcardially with 5–10 ml saline followed by 500 ml 4% paraformaldehyde in 0.1 m phosphate buffer. The brains were removed and immersed in the same fixative for 2 h. They were then washed overnight in 0.05 m Coon’s veronal buffer (pH 7.4) containing 20% sucrose, embedded in Tissue-Tek (Miles Laboratories, Naperville, CA), and frozen in liquid nitrogen. Frozen 14-␮m coronal sections were collected from the level of ⫹0.45 to ⫺0.46 mm relative to the bregma, according to the atlas of Swanson (11). The sections were mounted onto gelatin-coated slides and stored at ⫺80 C until used for in situ hybridization. All experiments were carried out in accordance with the European Communities Council Directive of November 24, 1986 (86/609/EEC), regarding mammalian research. 35

S-Labeled GAP-43 cRNA probes

The plasmid vector PGEM3Z containing a PstI/EcoRI fragment of 743 bp corresponding to bases 552-1295 of the rat GAP-43 complementary DNA (cDNA) (12) was used. PstI and T7 RNA polymerase were used to synthesize the antisense probe, and EcoRI with T3 RNA polymerase were used to synthesize the sense probe. The radioactive probes were generated in the presence of 200 ␮Ci [35S]CTP (Amersham Pharmacia Biotech, Les Ulis, France) using 1 ␮l of the appropriate RNA polymerase in a 40-␮l transcription reaction volume containing 8 ␮l 5 ⫻ transcription buffer, 2 ␮l 0.1 m dithiothreitol (DTT), 1 ␮g linearized plasmid (50 ng/ml), 1 ␮l Escherichia coli transfer RNA (5 mg/ml), 20 U RNasin, and 1 ␮l of a 10 mm stock solution of ATP, GTP, and UTP. The transcription reagents were incubated for 4 h at 39 C. Labeled probes were extracted with phenol-chloroform and purified on a Sephadex G-50 column. The 35 S-labeled riboprobes were diluted with hybridization buffer to a final concentration of 30,000 dpm/␮l.

Digoxigenin-labeled cRNA GnRH probe The plasmid vector GST7 containing the 330-bp BamHI/HindIII insert of GnRH cDNA was linearized with HindIII for antisense and with BamHI for sense probes. The riboprobes were synthesized in vitro with 1 ␮g linearized GnRH cDNA, 1 ⫻ digoxigenin RNA labeling mixture (Roche, Meylan, France), RNA polymerase (T7 for antisense and SP6 for sense), and 1 ⫻ transcription buffer. This mixture was incubated at 37 C (T7) or at 39 C (SP6) for 2 h. Residual DNA was digested with deoxyribonuclease. The probes were diluted 1:1000 with hybridization buffer.

Dual labeling in situ hybridization Sections were removed from storage at ⫺80 C, thawed, and placed into 0.1 m glycine-0.2 m Tris-HCl (pH 7.4) for 10 min before treatment with proteinase K (1 mg/ml in 100 mm Tris, pH 8.0, and 50 mm EDTA) for 15 min at 37 C. Slides were then immersed in 4% paraformaldehyde/0.1 m phosphate buffer for 15 min, and treated with 0.1 m triethanolamine (pH 8.0) for 10 min, followed by 0.25% acetic anhy-

FIG. 1. Darkfield microphotographs of hemisections at the level of the diagonal band of Broca (DBB) (a and b) and a line drawing (c) of the distribution of GAP-43 mRNA-expressing cells in the rostral forebrain. a, Control section incubated with 35S-labeled sense probe; b, section incubated with 35S-labeled antisense probe. GAP-43 mRNA is highly expressed in numerous cells of the DBB. MS, Medial septum. Magnification, ⫻100. dride for 10 min. The sections were dehydrated in increasing concentrations of ethanol and hybridized overnight at 55 C with a mixture of 35S- and digoxigenin-labeled cRNA that were diluted with 50% formamide, 10% dextran sulfate, 0.3 m NaCl, 20 mm Tris-HCl (pH 8.0), 5 mm EDTA, 1 ⫻ Denhardt’s solution, 0.5 mg/ml E. coli transfer RNA, 100 mm DTT, and 1% salmon sperm DNA. Slides were washed for 30 and 60 min in 4 ⫻ SSC (1 ⫻ SSC ⫽ 0.25 m NaCl and 0.01 m sodium citrate, pH 7.2) containing 10 mm DTT, followed by a 30-min wash in 0.3 m NaCl, 20 mm Tris-HCl (pH 8.0), 5 mm EDTA, and 50% formamide. Sections were then treated with ribonuclease A [20 ␮g/ml in 0.1 m Tris (pH 8.0), 0.5 m NaCl, and 0.5 m EDTA] for 30 min

1650


Endo • 2000 Vol 141 • No 5

FIG. 2. Darkfield microphotographs (a, c, and e) and line drawings (b, d, and f) of the distribution of GAP-43 mRNA-expressing cells in the rostral (a and b) and caudal (c and d) preoptic regions and in the supraoptic nucleus (SO; e and f). GAP-43 mRNA is highly expressed in the median preoptic nucleus (MEPO; a– d) and in the medial preoptic nucleus (MPN; c and d). Note that labeling is absent in the SO (e and f). aco, Anterior commissure; AVP, anteroventral preoptic nucleus; AVPV, anteroventral periventricular nucleus; MPO, medial preoptic area; och, optic chiasma; opt, optic tract; PVpo, preoptic periventricular nucleus; sch, suprachiasmatic nucleus. Magnification, ⫻100. at 37 C and rinsed in 2 ⫻ SSC for 15 min at 60 C and 0.1 ⫻ SSC for 15 min at 60 C. The sections were then washed in buffer 1 (100 mm Tris-HCl and 150 mm NaCl, pH 7.4) and incubated for 30 min in blocking buffer (1% Roche blocking agent in buffer 1). Sections were incubated for 4 h in buffer 1 containing antidigoxigenin Fab conju-

gated to alkaline phosphatase (Roche) which was diluted 1:250 with buffer 1 containing 1% normal sheep serum and 2.4 mg levamisole/10 ml. After rinsing for 10 min in buffer 1 and for 10 min in buffer 2 (100 mm Tris-HCl, 50 mm MgCl2, and 100 mm NaCl, pH 9.5), sections were incubated in chromogen solution (buffer 2 containing tetrazolium


1651

FIG. 3. Average density of grains per GAP-43 mRNA-expressing cell in the DBB (A), AVPV (B), MEPO (C), and MPN (D) during the different time points of the estrous cycle. The values are the mean ⫾ SEM. Significant differences (P ⬍ 0.05) among the average values for the different groups are noted as a vs. b and c vs. d. Di, Diestrous II; Pro, proestrus; Es, estrus. chloride and 5-bromo-4-chloro-3-indolyl phosphate). The reaction was stopped after 3 h by rinsing twice for 15 min each time in TE buffer (10 mm Tris-HCl, pH 8.0, and 1 mm EDTA). The slides were dehydrated in 70% ethanol containing ammonium acetate (300 mm) and 100% ethanol, and dipped in Ilford K5 emulsion (Saint-Priest, France). All sections were developed after a 20-day exposure.

Controls Specificity controls included incubation of the sections with 35S- and digoxigenin-labeled sense probe (Fig. 1a), pretreatment with ribonuclease, and coincubation with a 100-fold excess of unlabeled antisense probe. No labeling was observed on the control sections.

Quantitative analysis Sixteen sections of the diagonal band of Broca (DBB)/rostral preoptic area/medial preoptic area were analyzed per animal. During the first analysis, the grain density corresponding to GAP-43 mRNA levels in

non-GnRH cells was quantified by using the DensiRag computerized program of Biocom (Les Ulis, France). The boundaries of the hypothalamic nuclei studied were determined from observation of the corresponding azure blue-stained sections. Sections from DBB, anteroventral periventricular nucleus (AVPV), median preoptic nucleus (MEPO) and medial preoptic nucleus (MPN) were viewed under a ⫻60 epiillumination darkfield objective. Video images were obtained with a camera attached to the microscope. The OD over the third ventricle adjacent to the preoptic area and over the supraoptic nucleus that did not contain hybridization signals was measured, averaged, and used to calculate the mean background density, which was subtracted from the OD measurement of signals over the GAP-43 cells. Cells were identified as labeled with GAP-43 probe if the OD over the perikaryon was at least 5 times higher than the background. About 30 cells per each hypothalamic nuclei and per section were analyzed. During the second analysis, quantification of GAP-43 mRNA in GnRH neurons was studied. Firstly, GnRH mRNA-expressing cells were observed under a brightfield illumination, and GAP-43 mRNA-express-

1652


Endo • 2000 Vol 141 • No 5

FIG. 4. Simultaneous brightfield and epiillumination photomicrograph of the preoptic area of adult female rats, showing cells labeled with a digoxigeninconjugated cRNA probe for GnRH mRNA (darkly stained) and 35S-labeled cRNA probe for GAP-43 mRNA (white silver grains). The arrow indicates a GnRH neuron expressing GAP-43 mRNA; the arrowhead shows a single labeled GAP-43 mRNA-containing cell.

ing cells were observed under darkfield illumination. Thus, each digoxigenin-labeled GnRH neuron was examined for the presence of silver grains using alternative observations in bright- and darkfield illuminations. About 120 –160 GnRH cells were studied per animal. The total numbers of GnRH mRNA-expressing cells and of GnRH cells coexpressing GAP-43 mRNA per section were counted. From these data, the proportion of GnRH neurons expressing GAP-43 mRNA was calculated and averaged for each animal. Secondly, the grain density corresponding to GAP-43 mRNA levels was quantified in GnRH-expressing neurons isolated under brightfield illumination. The grain density overlying each cell was then visualized and quantified in the same manner as for GnRH-negative cells.

Statistical analysis The number of double labeled cells and the grain density per cell were averaged for each animal, and the mean ⫾ sem were calculated for each group. Differences among estrous cycle stages were assessed with oneway ANOVA followed by post-hoc Bonferroni’s t test to compare the percentage of GnRH neurons expressing GAP-43 mRNA. Differences between the groups were regarded as significant when P ⬍ 0.05. The same test was used to compare the levels of GAP-43 mRNA expression in GnRH neurons and non-GnRH cells.

Double immunohistochemistry procedure The immunohistochemical dual labeling procedure to identify GAP-43 and GnRH was based upon the description of Alonso et al. (13). Briefly, after perfusion fixation with 4% paraformaldehyde and 0.5% glutaraldehyde in 0.1 m phosphate buffer, pH 7.4, the brain was removed and postfixed for 48 h. Coronal 40-␮m Vibratome (Leica, Rueil Malmaison, France) sections were incubated for 48 h in mouse monoclonal anti-GAP-43 antibody (1:500; Sigma-Aldricht, St. Quentin Fallavier, France), and rabbit anti-GnRH antibody (1:2000; provided by Dr. G. Tramu, Bordeaux, France). After rinsing in PBS, the sections were incubated for 2 h in fluorescein-labeled donkey antimouse IgG and Cy5labeled donkey antirabbit IgG (1:200; The Jackson Laboratory, West Grove, PA). After rinsing in PBS, sections were mounted with Mowiol (Calbiochem, La Jolla, CA) and examined with a Leica Corp. confocal laser scanning microscope (Rockleigh, NJ). The characteristics of these antisera have been described by Meiri et al. (14) for GAP-43 and by Barry et al. (15) for GnRH.

RIAs Plasma LH levels were measured using materials supplied by the NIDDK rat pituitary hormone distribution program (Baltimore, MD), and values were expressed as nanograms per ml using the LH pituitary reference preparation RP-3. Assay sensitivity was 0.02 ng/tube, and intra- and interassay variances were 6% and 8.5%, respectively. Plasma estradiol was measured using a RIA kit optimized for the direct quantitative determination of very low concentrations of 17␤estradiol in human serum plasma (e.g. in children), purchased from SORIN Biomedica (Antony, France). Assay sensitivity was 0.2 pg/tube, and intra- and interassay variances were 5.6% and 7.3%, respectively. Progesterone levels were measured in plasma samples without extraction, using a RIA kit purchased from SORIN Biomedica. Assay sensitivity was 5 pg/tube, and intra- and interassay variances were 5.5% and 8.1%, respectively.

Results Distribution of GAP-43 mRNA-expressing cells

High levels of GAP-43 mRNA hybridization signal were observed in certain cells of the DBB (Fig. 1, b and c) and in several nuclei of the rostral and medial preoptic regions, e.g. the AVPV, MEPO, and MPN (Fig. 2, a and c). GAP-43 mRNA levels were very low in the supraoptic nucleus (Fig. 2e). Cellular labeling intensities for GAP-43 mRNA during the estrous cycle

The average GAP-43 labeling density was calculated over cells in the DBB, AVPV, MEPO, and MPN at each of the six time points. Figure 3 depicts average brightness data from the different nuclei. Statistical comparisons revealed that the GAP-43 mRNA levels were higher in the DBB on the day of proestrus than on diestrous II (P ⬍ 0.05; Fig. 3A), whereas GAP-43 mRNA levels did not vary throughout the estrous cycle in the AVPV (Fig. 3B). In the MEPO, GAP-43 mRNA levels were slightly higher in diestrus and proestrus than in estrus (Fig. 3C), whereas in the MPN, GAP-43 mRNA levels


were elevated only at 1600 h on proestrus compared with those at 1600 h on estrus (P ⬍ 0.05; Fig. 3D).

1653

GAP-43 mRNA was expressed in numerous GnRH neurons (Fig. 4) without preferential distribution according to the different areas.

Distribution of GAP-43 mRNA-expressing GnRH neurons

The distribution of neurons expressing GnRH mRNA was similar to the description of previous in situ hybridization studies (16 –19). The greatest numbers of labeled GnRH neurons were seen in the rostral preoptic area, followed in decreasing order by the medial preoptic area and the DBB.

FIG. 5. Percentage of GnRH neurons expressing GAP-43 mRNA (A) and relative amount of GAP-43 mRNA in GnRH neurons (B) throughout the rat estrous cycle. The values are the mean ⫾ SEM. The number of GnRH neurons that express GAP-43 mRNA is highest during proestrus, as is the relative amount of GAP-43 cRNA hybridization signal per GnRH neuron. Significant differences (P ⬍ 0.05) among the average values for the different time points are indicated as a vs. b and c vs. d.

Changes in the number of GnRH cells expressing GAP-43 mRNA during the estrous cycle

GAP-43 mRNA was expressed in many GnRH neurons throughout the estrous cycle; however, the percentage of dual labeled neurons changed significantly depending upon

1654


Endo • 2000 Vol 141 • No 5

the stage of the cycle (Fig. 5A). The percentage of GnRH neurons that expressed GAP-43 mRNA was lowest during diestrous II (40 ⫾ 2%; P ⬍ 0.001); it peaked during proestrus (64 ⫾ 5%) and returned to basal on estrus (45 ⫾ 8%; P ⬍ 0.05). Changes in the GAP-43 mRNA labeling intensities in the GnRH neurons during the estrous cycle

The average GAP-43 labeling density per GnRH cell was 29 ⫾ 1 when all animal groups were averaged. Dual labeled GnRH neurons generally expressed lower GAP-43 mRNA than single labeled GAP-43 cells. Figure 5B lists the average density of grains per GnRH neuron during different stages of the estrous cycle. Statistical comparisons revealed differences between several time points. GAP-43 mRNA hybridization signal in GnRH cells was low at 1600 h on diestrous II; it was increased significantly by 0800 h on proestrus and was maintained at high levels through 1600 h on estrus (Fig. 6). When the morning and afternoon data were grouped, GAP-43 mRNA levels were elevated on proestrus (density of grains per cell, 29 ⫾ 1) and on estrus (density of grains per cell, 32 ⫾ 2) compared with those on diestrous II (density of grains per cell, 25 ⫾ 1; P ⬍ 0.01; Fig. 5B). GAP-43 protein is expressed in GnRH nerve terminals in the median eminence of adult animals

In the median eminence, GAP-43 labeling was observed in the same regions where most GnRH axons and terminals were located. Dual immunohistochemistry for GnRH and GAP-43 showed that numerous GnRH-positive axons and nerve endings were also GAP-43 positive (Fig. 7). In addition, many GAP-43-positive elements did not contain GnRH (Fig. 7). Periovulatory hormone secretions

Average plasma LH, estrogen, and progesterone levels were typical for the different time points of the estrous cycle (Table 1). Thus, LH levels were high only during the preovulatory surge at 1600 h on proestrus, and estrogen showed the expected high values during proestrus and had returned to baseline values by 0800 –1600 h on estrus. Average progesterone levels reached two peaks during the estrous cycle. The first peak was observed at 0800 h on diestrous II, and the second peak began during the afternoon of proestrus. These results are in accordance with our previous results (5) and those of Butcher et al. (10). Discussion

The results of the present study show that GAP-43 mRNA and protein are expressed in adult GnRH neurons and that the percentage of GnRH neurons expressing GAP-43 changes FIG. 6. Brightfield photomicrographs showing GnRH neurons (darkly stained) expressing GAP-43 mRNA (black silver grains) in diestrus (a), proestrus (b), and estrus (c). Note the lower level of expression of GAP-43 mRNA in GnRH cells (arrowhead) on the day of diestrus compared with that on the day of proestrus. In d, the GnRH neuron (empty arrow) does not express GAP-43 mRNA. The arrows show GAP-43 mRNA-expressing cells that do not contain GnRH mRNA. Magnification, ⫻1000.


1655

FIG. 7. Confocal images of 40-␮m thick frontal section through the median eminence. Double immunostaining for GAP-43 (left panel) and GnRH (right panel) shows that the two immunostainings are colocalized within certain axons and nerve terminals (arrows) located in the external layer of the median eminence. Magnification, ⫻400. TABLE 1. Average ⫾

SEM

concentrations of plasma LH, estradiol, and progesterone in rats killed at different estrous cycle stages

Diestrous II, 0800 h Diestrous II, 1600 h Proestrus, 0800 h Proestrus, 1600 h Estrus, 0800 h Estrus, 1600 h a b

LH (ng/ml)

Estradiol (pg/ml)

Progesterone (ng/ml)

0.59 ⫾ 0.17 0.49 ⫾ 0.1 1.26 ⫾ 0.31 4.24 ⫾ 1.57a 0.49 ⫾ 0.12 0.38 ⫾ 0.08

11.28 ⫾ 2.38 8.36 ⫾ 2.84 31.95 ⫾ 6.20a 62.48 ⫾ 9.51b 3.15 ⫾ 0.25 4.98 ⫾ 2.61

12.53 ⫾ 1.81 12.88 ⫾ 1.05 11.08 ⫾ 1.33 23.24 ⫾ 5.31a 15.60 ⫾ 1.72c 15.45 ⫾ 1.84c

Significantly different from the other estrous stages (P ⬍ 0.05), except from c. Significantly different from a and the other estrous cycle stages (P ⬍ 0.05).

during the rat estrous cycle. This percentage is the highest during the day of the preovulatory LH surge. Similarly, GAP-43 mRNA content in individual GnRH neurons varies during estrous cycle. Thus, GAP-43 mRNA levels in GnRH neurons are the lowest during diestrous II, and they increase by 50% during proestrus. These high levels of GAP-43 mRNA are maintained for at least 36 h (until 1600 h of estrus). Together, these results suggest that GnRH neurons show an increase in GAP-43 mRNA levels just before the preovulatory gonadotropin surge, i.e. in proestrus. It is well established that the proestrous surge of GnRH (1, 2) provides the neural trigger for the release of the preovulatory gonadotropin surge that, in turn, induces ovulation (20). Although it is known that GnRH neurons are acutely activated under the preovulatory hormonal conditions, i.e. by high estradiol levels during the morning and afternoon of proestrus (5, 10) and by rising progesterone levels during the afternoon of proestrus (21, 22), the mechanisms by which gonadal and circadian (23) cues are conveyed to the GnRH neurons remain unclear. It appears that both the GnRH perikaryon as well as the axon terminals undergo specific changes in response to elevated estradiol levels. Thus, we recently reported that physical contacts between GnRH nerve endings and the perivascular space in the external zone of the median eminence were observed only on

the day of proestrus, whereas during the other stages of the cycle, glial elements prevented direct access of GnRH terminals to the capillaries (6). This neuro-glio-endothelial plasticity allows direct access of the GnRH nerve terminals to the fenestrated capillaries only during the time of high secretory activity during which the GnRH nerve terminals project filipodia-like extensions toward the basal lamina (5, 6). The results of the present study suggest that the outgrowth of GnRH nerve terminals on the day of proestrus is stimulated by GAP-43, which is an important regulator of neurite outgrowth in many neuronal systems. Indeed, both the percentage of GnRH cells expressing GAP-43 mRNA and the GAP-43 mRNA levels in GnRH cells are highest at this estrous cycle stage. As GAP-43 protein levels appear to be proportional to the mRNA levels (12, 24 –29), it is suggested that increased expression of the GAP-43 mRNA in GnRH neurons leads to a rapid accumulation of GAP-43 protein in the GnRH axon terminals in the external zone of the median eminence (30 –32). The presence of GAP-43 protein in this brain region has been reported (13), and the results of the present study show that this protein is colocalized with many GnRH nerve fibers and terminals. GAP-43 could cause an enhanced signal for growth (7, 8), such that GnRH neuroendocrine axons sprout toward the basal lamina, which would facilitate GnRH release into the portal blood. This scenario is

1656


similar to that for neurons of transgenic mice, in which an enhanced expression of GAP-43 induces and potentiates nerve sprouting in the adult animal by promoting growth cone guidance and the formation and growth of sprouts (7). It is likely that circulating gonadal steroids provide important endocrine cues for the regulation of GAP-43 gene expression in GnRH neurons and also in many surrounding cells. Thus, estrogen exerts dramatic effects on the growth and phenotype of basal forebrain neurons, particularly in the preoptic area and hypothalamus (33) where estrogen receptor-containing cells are abundant (34). The trophic actions of estrogen include enhancement of neurite outgrowth (35), dendritic spine formation (36), and synaptic density (37). Previous in situ hybridization studies have shown that treatment of ovariectomized adult rats with estrogen elevates GAP-43 mRNA in the medial preoptic area and hypothalamus (38 – 40). Shughrue and Dorsa (38) showed that GAP-43 mRNA levels in the medial preoptic area were elevated in estrogen-treated ovariectomized animals compared with those in ovariectomized animals. Our high resolution autoradiograms indicate that under physiological conditions, such as during the estrous cycle, the GAP-43 mRNA levels in non-GnRH cells of the preoptic area do not vary as markedly, whereas greater fluctuations in GAP-43 mRNA levels were observed in the DBB. In most nuclei analyzed in this study, lower cellular levels of GAP-43 mRNA were observed during the stages of the estrous cycle that are characterized by low plasma estrogen levels. However, no variation of GAP-43 gene expression was observed in the AVPV, which is a critical brain region where estrogen is thought to act as an inducer of the GnRH/LH surge (34). With respect to the GnRH neurons themselves, although an action of progesterone cannot be excluded, it is suggested that the dramatic increase in plasma estrogen levels between diestrus and proestrus is responsible for the elevation of GAP-43 mRNA levels and the percentage of GnRH neurons expressing GAP-43 mRNA. As GnRH neurons do not express estrogen receptors (41), estrogen could regulate GAP-43 gene expression in these neurons indirectly through stimulation of estradiol target neurons, which, in turn, convey the steroid signal to the GnRH neurons (34). In conclusion, the results of this study indicate that GAP-43 is expressed in adult GnRH neurons and that the GAP-43 mRNA expression rate varies during the rat estrous cycle. Elevated levels of GAP-43 mRNA in GnRH neurons on the day of proestrus may lead to an accumulation of GAP-43 protein in GnRH nerve terminals at the median eminence. This phenomenon could contribute to an enhanced responsiveness of GnRH nerve endings to growth factors produced locally in the median eminence and promote sprouting of GnRH nerve terminals toward the pericapillary space. This would permit neurovascular contacts and thus facilitate GnRH release into the pituitary portal blood. Acknowledgments The authors thank Mrs. G. Mortreux for her excellent technical assistance with RIA realization, and the NIDDK for the rat LH RIA material. We also thank Mrs. R. Bogaert and Mrs. Christine Declerck (Leica Corp. Microsystem SA, Rueil-Malmaison, France) for assistance with making the figures.

Endo • 2000 Vol 141 • No 5

References 1. Sarkar DK, Chiappa SA, Fink G, Sherwood NM 1976 Gonadotropin-releasing hormone surge in pro-oestrous rats. Nature 264:461– 463 2. Levine JE, Ramirez VD 1982 Luteinizing hormone-releasing hormone release during the rat estrous cycle and after ovariectomy, as estimated with push-pull cannulae. Endocrinology 111:1439 –1448 3. Lee WS, Abbud R, Smith MS, Hoffman GE 1992 LHRH neurons express c-Jun protein during the proestrus surge of luteinizing hormone. Endocrinology 130:3101–3103 4. Lee WS, Smith MS, Hoffman GE 1990 Luteinizing hormone-releasing hormone neurons express Fos protein during the proestrus surge of luteinizing hormone. Proc Natl Acad Sci USA 87:5163–5167 5. Prevot V, Dutoit S, Croix D, Tramu G, Beauvillain JC 1998 Semi-quantitative ultrastructural analysis of the localization and neuropeptide content of gonadotropin-releasing hormone nerve terminals in the median eminence throughout the estrous cycle of the rat. Neuroscience 84:177–191 6. Prevot V, Croix D, Bouret S, Dutoit S, Tramu G, Stefano S, Beauvillain JC 1999 Definitive evidence for the existence of morphological plasticity in the external zone of the median eminence during the rat estrous cycle: implication of neuro-glio-endothelial interactions in the gonadotropin-releasing hormone release. Neuroscience 94:809 – 819 7. Aigner L, Arber S, Kapfhammer JP, Laux T, Schneider C, Botteri F, Brenner HR Caroni P 1995 Overexpression of the neural growth-associated protein GAP-43 induces nerve sprouting in the adult nervous system of transgenic mice. Cell 83:269 –278 8. Benowitz LI, Routtenberg A 1997 GAP-43:an intrinsic determinant of neuronal development and plasticity. Trends Neurosci 20:84 –91 9. Sarkar DK, Minami S 1995 Diurnal variation in luteinizing hormone-releasing hormone and ␤-endorphin release in pituitary portal plasma during the rat estrous cycle. Biol Reprod 53:38 – 45 10. Butcher RL, Collins WE, Fugo NW 1974 Plasma concentration of LH, FSH, prolactin, progesterone and estradiol-17␤ throughout the 4-day estrous cycle of the rat. Endocrinology 94:1704 –1708 11. Swanson LW 1992 Structure of the Rat Brain. Elsevier, Amsterdam 12. Meberg PJ, Routtenberg A 1991 Selective expression of protein F1/(GAP-43) mRNA in pyramidal but not granule cells of the hippocampus. Neuroscience 45:721–733 13. Alonso G, Prieto M, Legrand A, Chauvet N 1997 PSA-NCAM and B-50/ GAP-43 are coexpressed by specific neuronal systems of the adult rat mediobasal hypothalamus that exhibit remarkable capacities for morphological plasticity. J Comp Neurol 384:181–199 14. Meiri KF, Bickerstaff LE, Schwob JE 1991 Monoclonal antibodies show that kinase C phosphorylation of GAP-43 during axonogenesis is both spatially and temporally restricted in vivo. J Cell Biol 112:991–1005 15. Barry J, Dubois MP, Carette B 1974 Immunofluorescence study of the preoptico-infundibular LRF neurosecretory pathway in the normal, castrated or testosterone-treated male guinea pig. Endocrinology 95:1416 –1423 16. Porkka-Heiskanen T, Urban JH, Turek FW, Levine JE 1994 Gene expression in a subpopulation of luteinizing hormone-releasing hormone (LHRH) neurons prior to the preovulatory gonadotropin surge. J Neurosci 14:5548 –5558 17. Ronnekleiv OK, Naylor BR, Bond CT, Adelman JP 1989 Combined immunohistochemistry for gonadotropin-releasing hormone(GnRH) and proGnRH, and in situ hybridization for GnRH messenger ribonucleic acid in rat brain. Mol Endocrinol 3:363–371 18. Malik KF, Silverman AJ, Morrell JI 1991 Gonadotropin-releasing hormone mRNA in the rat: distribution and neuronal content over the estrous cycle and after castration of males. Anat Rec 231:457– 466 19. Mitchell V, Prevot V, Jennes L, Aubert JP, Croix D, Beauvillain JC 1997 Presence of ␮ and ␬ opioid receptor mRNAs in galanin but not in GnRH neurons in the female rat. Neuroreport 8:3167–3172 20. Schwartz NB 1974 The role of FSH and LH and of their antibodies on follicle growth and on ovulation. Biol Reprod 10:236 –272 21. Krey LC, Tyrey L, Everett JW 1973 The estrogen-induced advance in the cyclic LH surge in the rat: dependency on ovarian progesterone secretion. Endocrinology 93:385–390 22. Feder HH, Brown-Grant K, Corker CS 1971 Pre-ovulatory progesterone, the adrenal cortex and the ‘critical period’ for luteinizing hormone release in rats. J Endocrinol 50:29 –39 23. Legan SJ, Karsch FJ 1975 A daily signal for the LH surge in the rat. Endocrinology 96:57– 62 24. Basi GS, Jacobson RD, Virag I, Schilling J, Skene JH 1987 Primary structure and transcriptional regulation of GAP-43, a protein associated with nerve growth. Cell 49:785–791 25. Karns LR, Ng SC, Freeman JA, Fishman MC 1987 Cloning of complementary DNA for GAP-43, a neuronal growth-related protein. Science 236:597– 600 26. Ng SC, de la Monte SM, Conboy GL, Karns LR, Fishman MC 1988 Cloning of human GAP-43: growth association and ischemic resurgence. Neuron 1:133–139 27. Rosenthal A, Chan SY, Henzel W, Haskell C, Kuang WJ, Chen E, Wilcox JN, Ullrich A, Goeddel DV, Routtenberg A 1987 Primary structure and mRNA


28.

29.

30. 31.

32.

33.

localization of protein F1, a growth-related protein kinase C substrate associated with synaptic plasticity. EMBO J 6:3641–3646 Hoffman PN 1989 Expression of GAP-43, a rapidly transported growthassociated protein, and class II ␤ tubulin, a slowly transported cytoskeletal protein, are coordinated in regenerating neurons. J Neurosci 9:893– 897 Van der Zee CE, Nielander HB, Vos JP, Lopes da Silva S, Verhaagen J, Oestreicher AB, Schrama LH, Schotman P, Gispen WH 1989 Expression of growth-associated protein B-50 (GAP43) in dorsal root ganglia and sciatic nerve during regenerative sprouting. J Neurosci 9:3505–3512 Skene JH 1989 Axonal growth-associated proteins. Annu Rev Neurosci 12:127–156 Gispen WH, Leunissen JL, Oestreicher AB, Verkleij AJ, Zwiers H 1985 Presynaptic localization of B-50 phosphoprotein: the (ACTH)-sensitive protein kinase substrate involved in rat brain polyphosphoinositide metabolism. Brain Res 328:381–385 DiFiglia M, Roberts RC, Benowitz LI 1990 Immunoreactive GAP-43 in the neuropil of adult rat neostriatum: localization in unmyelinated fibers, axon terminals, and dendritic spines. J Comp Neurol 302:992–1001 Garcia-Segura LM, Chowen JA, Parducz A, Naftolin 1994 Gonadal hormones as promoters of structural synaptic plasticity: cellular mechanisms. Prog Neurobiol 44:279 –307

1657

34. Herbison AE 1998 Multimodal influence of estrogen upon gonadotropinreleasing hormone neurons. Endocr Rev 19:302–330 35. Ferreira A, Caceres A 1991 Estrogen-enhanced neurite growth: evidence for a selective induction of Tau and stable microtubules. J Neurosci 11:392– 400 36. Gould E, Woolley CS, Frankfurt M, McEwen BS 1990 Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood. J Neurosci 10:1286 –1291 37. Woolley CS, McEwen BS 1992 Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat. J Neurosci 12:2549 –2554 38. Shughrue PJ, Dorsa DM 1993 Estrogen modulates the growth-associated protein GAP-43 (neuromodulin) mRNA in the rat preoptic area and basal hypothalamus. Neuroendocrinology 57:439 – 447 39. Lustig RH, Sudol M, Pfaff DW, Federoff HJ 1991 Estrogenic regulation and sex dimorphism of growth-associated protein 43 kDa (GAP-43) messenger RNA in the rat. Brain Res Mol Brain Res 11:125–132 40. Singer CA, Pang PA, Dobie DJ, Dorsa DM 1996 Estrogen increases GAP-43 (neuromodulin) mRNA in the preoptic area of aged rats. Neurobiol Aging 17:661– 663 41. Shivers BD, Harlan RE, Morrell JI, Pfaff DW 1983 Absence of oestradiol concentration in cell nuclei of LHRH-immunoreactive neurones. Nature 304: 345–347