0013-7227/90/1271-0365$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 127, No. 1 Printed in U.S.A.
Gonadotropin-Releasing Hormone Gene Expression during the Rat Estrous Cycle: Effects of Pentobarbital and Ovarian Steroids* OK-KYONG PARK, SAJIV GUGNEJA, AND KELLY E. MAYO Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, Euanston, Illinois 60208
ABSTRACT. Although hypothalamic GnRH release is known to be modulated by neural and hormonal factors, the relationship between altered GnRH secretion and GnRH synthesis remains unclear. In an attempt to address this question, we examined GnRH gene expression in the rat hypothalamus using in situ hybridization histochemistry. An 25S-labeled antisense RNA probe was used to identify neurons expressing GnRH mRNA in an area that included the diagonal band of Broca, the organum vasculosum of the lamina terminalis, and the preoptic area. The number of GnRH mRNA-expressing cells was determined at various times during the rat estrous cycle. During proestrus, the number of GnRH mRNA-expressing cells decreased somewhat at 1400-1600 h, increased significantly at 1800 h (the time of the LH surge), then gradually returned to basal levels at 2200 h. Expression did not change substantially at other times during the estrous cycle. To understand this close temporal relationship between the LH surge and increased GnRH mRNA levels, we examined GnRH gene expression in proestrous animals in which
the LH surge was blocked with pentobarbital. Pentobarbital treatment blocked the increase in the number of GnRH mRNAexpressing cells normally observed at 1800 h in saline-treated controls, suggesting that the increase in GnRH gene expression is closely coupled to secretion of GnRH from the hypothalamus. Finally, we addressed the question of whether ovarian steroids have direct effects on GnRH gene expression by examining GnRH mRNA levels in ovariectomized steroid-treated rats at a time before (1100 h) and a time after (1800 h) hypothalamic GnRH hypersecretion. At 1100 h, no significant changes were observed, but at 1800 h, estrogen-treated rats showed a significant increase in both the number of GnRH mRNA-expressing cells and serum LH levels. This suggests that estrogen influences GnRH gene expression indirectly, perhaps by altering hypothalamic GnRH release. Our results in each of these models suggest that GnRH mRNA levels increase in response to GnRH hypersecretion at the time of the LH surge. (Endocrinology 127: 365372,1990)
G
in profound alterations in the estrous cycle (8). It also results in topological changes in GnRH content, such as elevation in the area anterior to the optic chiasm (9) and severe depletion in the median eminence (10). After the isolation of GnRH cDNA clones and the prediction of the GnRH precursor sequence (11, 12), several groups have used in situ hybridization, Northern or dot blot RNA analysis, and solution hybridization to determine changes in GnRH gene expression in animals treated with different hormones. These results, however, have been difficult to interpret. For example, estrogen, which induces the LH surge, has been reported to exert stimulatory (13-15), inhibitory (16-18), or no (19) effects on GnRH gene expression. In addition, elimination of gonadal steroid feedback by ovariectomy (OVX) is reported to have diverse effects on GnRH gene expression (14, 18, 20). These differing results might be attributable to differences in the sensitivity of techniques, anatomical regions analyzed, and/or time points used in these studies. To understand dynamic changes in GnRH gene expression in more detail, we attempted to examine GnRH gene
nRH NEURONS originate from the medial offactory placode during embryonic development (1) and migrate predominantly into the diagonal band of Broca (DBB), the organum vasculosum of the lamina terminalis (OVLT), and the preoptic area (POA) (2). GnRH cells in the OVLT-POA-DBB region are believed to be directly involved in controlling the function of pituitary gonadotropes, since a majority of GnRH cells from these areas send projections to the median eminence (3) and secrete GnRH into the hypophysial portal bloodstream. GnRH activates pituitary GnRH receptors (4) and thereby modulates the synthesis and/or release of gonadotropins (5-7). GnRH cells in the OVLT-POA area are clearly important for normal reproductive function, since deafferentation of the hypothalamus results Received February 14, 1990. Address all correspondence and requests for reprints to: Dr. Kelly E. Mayo, Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, 2153 Sheridan Road, Evanston, Illinois 60208. *This work was supported by a development award from the McKnight Fund for Neuroscience and NIH Grants NS-24439 and HD21921.
365
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
366
expression at many time points in intact rats during the estrous cycle, in proestrous rats treated with pentobarbital, and in OVX animals treated with combinations of ovarian steroids. A sensitive and specific in situ hybridization assay was used to determine the number of GnRH mRNA-expressing neurons in the POA-OVLT-DBB area, while serum LH levels were used as an index of hypothalamic GnRH release. Materials and Methods Animals Intact Sprague-Dawley female rats (200-240 g BW) were purchased from Charles River Breeding Laboratories (Wilmington, MA) and kept in a photoperiod of 14 h of light and 10 h of darkness, with lights on at 0500 h. Food and water were available ad libitum. Stages of the estrous cycle of a group of rats were monitored by daily vaginal lavage, and only those rats demonstrating at least two consecutive 4-day cycles were used in this study. Another group of rats was bilaterally OVX under methoxyflurane anesthesia. Three different experimental groups were included in this study: intact rats during various stages of the estrous cycle, pentobarbital-treated proestrous animals, and steroid-replaced OVX animals. Estrous cycle experiment. Rats (n = 4 or 5 for each time point) were killed at different time intervals (at 1100 and 1800 h on estrus, metestrus, and diestrus and at 0900, 1100, 1400, 1600, 1800, 2000, and 2200 h on proestrus). After decapitation, trunk blood was collected for subsequent serum LH determination, and the brain was rapidly removed, frozen on dry ice, and stored at -70 C until used. Pentobarbital experiment. At 1100 h on proestrus, rats (n = 4 in each group) were injected ip with either sodium pentobarbital (100 mg/kg BW) or saline. At 1800 h, rats were decapitated, and brains and trunk blood were collected as described above. Steroid-replacement experiment. Three groups of rats were used for this experiment: OVX, OVX estrogen-primed (OVX + E2), and OVX estrogen- plus progesterone-primed (OVX + E2 + P4) rats. Each of these groups was examined at the time points of 1100 and 1800 h (n = 4 in each group). At 0930-1000 h 7-9 days after OVX surgery, animals were implanted with either a crystalline E2 capsule (5 mm; Dow-Corning Silastic tubing, Medfield, MA; id, 0.062 in.; od, 0.095 in.) or an empty capsule. At 0900 h 2 days later, animals were sc injected with either progesterone (1.5 mg in mineral oil) or vehicle only. Rats were decapitated at either 1100 or 1800 h, and brains and trunk blood were collected as described above.
NaCl, 30 mM sodium citrate, pH 7.0) for 5 min, dipped in distilled deionized H2O and 0.1 M triethanolamine, and incubated in 0.25% acetic anhydride in 0.1 M triethanolamine for 10 min. Sections were dehydrated through an ethanol series and vacuum dried until hybridization. The GnRH clone was isolated from a rat hypothalamic cDNA library by screening with an oligonucleotide corresponding to the mature GnRH decapeptide. The largest clone isolated contained a 575-basepair (bp) insert that includes 128 bp of 5' nontranslated sequences, the complete coding region for the 92-amino acid GnRH precursor, 146 bp of 3' nontranslated sequennes, and a 25-bp poly (A) tract. The sequence of the clone corresponds to that previously reported by others (11, 12) for the rat GnRH cDNA and gene. The cDNA insert was subcloned in both orientations into pGEM-4 for probe preparation (see Fig. 1). Antisense and sense [35S]UTP-labeled RNA probes were synthesized using SP6 polymerase (22) from GnRH-C digested with HindI.II (Fig. 1A) and GnRH-A digested with Kpnl (Fig. IB), respectively. Probe (107 cpm/ml hybridization buffer) was applied to the slides, and they were coverslipped and hybridized at 47 C for 15 h. After hybridization, the coverslips were removed in 4 x SSC, and sections were treated with RNase-A (20 ng/va\) at 37 C for 30 min, washed to a stringency of 0.25 x SSC at 55 C, and dehydrated through an ethanol series. Sections were exposed on Kodak XAR-5 film (Eastman Kodak, Rochester, NY) overnight, and then dipped in Kodak NTB-2 liquid emulsion and exposed for 2 weeks. After developing and fixing, sections were stained with cresyl violet, observed, and photographed using a Nikon (Garden City, NY) optiphot microscope. The microscope was equipped with a Javelin (Torrance, CA) solid state camera and Sony (Park Ridge, NJ) monitor to facilitate cell counting.
LHRIA LH concentrations were determined using an ovine:rat RIA system employing a NIDDK kit (NIH LH S-25 as standard
T7
100bp
Antisense RNA Probe
b
GnRH in situ hybridization Coronal sections (20 j^m) of the brain throughout the DBB, OVLT, and POA were cut on a Reichert 820 cryostat (Reichert, Buffalo, NY) at -20 C and mounted onto gelatin and poly-Llysine-coated glass slides for in situ hybridization, as described previously (21). Briefly, brain sections were fixed in 5% paraformaldehyde (pH 7.5) for 5 min, washed in 2 x SSC (300 mM
Endo • 1990 Vol 127 • No 1
t> * if
1— PGEM4 100 bp
Sense RNA Probe FIG. 1. Schematic structure of the rat GnRH cDNA clone used in this study. A full-length GnRH cDNA was subcloned into the vector pGEM4 either in an antisense (A) or a sense (B) orientation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE and antirat LH antibody S-10). Results are expressed as nanograms per milliliter NIH LH S-25 standard. The intra- and interassay coefficients of variation were 5.0% and 6.0%, respectively.
100X:OVLT/POA
367 500X: OVLT/POA
Data analysis GnRH mRNA-expressing cells were manually counted in 44-46 sections (20 /*m) from an area encompassing the DBB, OVLT, and POA. The average number of GnRH mRNAexpressing cells per brain section in the POA-OVLT-DBB area (GN/S) was determined for each animal. The number of GnRH mRNA-expressing cells in cycling animals from five independent experiments was expressed as a percentage of the value at 1800 h on metestrus in each experiment. The number of hybridizing cells per section at 1800 h on metestrus ranged between 19.9-22.4 in the five experiments (mean, 20.9 ± 0.8). For the pentobarbital and steroid replacement experiments, the values at 1800 h in saline-treated and OVX + E2 rats were considered as 100%. GN/S ranged between 22.9-33.3 at 1800 h on proestrus in saline-treated animals (mean, 28.2 ± 3.2) and between 23.0-28.9 at 1800 h in OVX + E2 animals (mean, 25.4 ± 1.2). One-way analysis of variance followed by the post-hoc test were used for statistical analysis of data from the cycling and steroid replacement experiments, while the group Student t test was used for data from the pentobarbital experiment. P < 0.05 was considered significant.
Results Figure 1 illustrates the structure of a full-length GnRH cDNA clone that was isolated from a rat hypothalamic cDNA library and subcloned into the plasmid vector pGEM-4 in either an antisense (Fig. 1A) or a sense (Fig. IB) orientation. This clone was used to synthesize riboprobes for in situ hybridization histochemistry in the rat brain. Examination of brain sections with an [35S]UTPlabeled antisense RNA probe revealed hybridizing cells in the area of the DBB, OVLT, and POA, areas known as loci for GnRH-immunoreactive cells. The hybridizing cells had a random morphology typical of GnRH cells: oval, fusiform, and spindly shaped. Figure 2 shows an example of brain sections containing GnRH mRNAexpressing cells. Silver grains indicative of positive hybridization appear to be deposited over individual cells (Fig. 2, A and C) and are easily discernable over background, particularly using darkfield microscopy (Fig. 2, B and D). The specificity of hybridization was tested in sequential brain sections that were taken at the level of the OVLT-POA and hybridized with either an antisense or a sense strand GnRH riboprobe. Figure 3A shows several dozen cells expressing GnRH mRNA observed using the antisense probe; no hybridization was observed using the sense probe (Fig. 3B). In addition, the GnRH antisense probe did not hybridize to a variety of control tissues examined, including liver and ovary (data not shown).
FIG. 2. Examples of GnRH cells in the POA after hybridization to a GnRH antisense RNA probe. A and B are photographs taken at X100 magnification; C and D are photographs taken at X500 magnification. The upper panels were photographed using brightfield optics, and the lower panels used darkfield optics.
40X: OVLT/POA
o C Q) CO
o ol a> (/> c a> (A
FIG. 3. Darkfield microscopy of adjacent brain sections at the level of the OVLT-POA. Sections were hybridized with either an [35S]UTPlabeled antisense RNA or an [35S]UTP-labeled sense RNA probe. Photographs were taken at X40 magnification.
To determine whether GnRH gene expression changes during the estrous cycle, we examined GnRH mRNA levels in animals at different times throughout the 4-day cycle. As an index of GnRH mRNA expression, we counted the number of hybridizing cells in an area that included the DBB, OVLT, and POA. We also measured serum LH levels in the animals used for this study. These
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
368
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
Endo • 1990 Vol 127 • No 1
200-1
o o CO CO
•8 1 0 ° H FIG. 4. Changes in the number of GnRH mRNA-expressing cells and in serum LH levels during the rat estrous cycle (n = 4 or 5 for each group). A, GN/S is the average number of GnRH mRNA-positive cells per brain section; values are expressed as a percentage of GN/S at 1800 h on metestrus. B shows serum LH values in the corresponding animals. Asterisks indicate P < 0.05 compared to the other time points.
z o
B c 10-
1200
1200
Metestrus
data are presented in Fig. 4. Figure 4A shows that the number of GnRH mRNA-expressing cells, expressed as a percentage of the value at 1800 h on metestrus, is maintained at a fairly constant level throughout the estrous cycle, except during proestrus. During proestrus, the number of GnRH mRNA-expressing cells decreases somewhat at 1400-1600 h, then increases about 2-fold (P < 0.01) at 1800 h, and gradually returns to basal levels at 2200 h. Serum LH levels in these animals were significantly elevated by 1600 h on proestrus, 2 h before the rise in the number of GnRH mRNA-expressing cells, and remained high throughout the evening of proestrus (Fig. 4B). The close temporal correlation between the increased number of GnRH mRNA-expressing cells and the onset of the preovulatory LH surge prompted us to further investigate the relationship between these events. We used pentobarbital to eliminate the LH surge in proestrous rats and examined GnRH mRNA levels in these animals. Figure 5 shows the effects of pentobarbital on serum LH levels and GnRH gene expression. As expected, pentobarbital treatment completely blocked the preovulatory LH surge (P < 0.01), which remained unaffected in saline-treated animals. The number of GnRH mRNA-expressing cells was also significantly lower (P
1200
Diestrus Proestrus Rat Colony Time (h)
1200
Estrus
< 0.01) in pentobarbital-treated rats compared to that in saline-treated animals (65.1 ± 7.6% us. 100.0 ± 4.3%; Fig. 5). Finally, we examined the effect of ovarian steroid feedback on GnRH gene expression. Animals were OVX for 7-9 days, then replaced with either estrogen alone or estrogen plus progesterone. Rats were subsequently examined at two times of day: 1100 h, a time when the number of GnRH mRNA-expressing cells has not significantly changed in proestrous animals, and 1800 h, a time when the number of GnRH-mRNA-expressing cells is maximal in proestrous animals (Fig. 4). As shown in Fig. 6, no differences were seen among the three groups at 1100 h. However, a small but significant (P < 0.05) increase in the number of GnRH mRNA-expressing cells was observed in the OVX + E2 group. Serum LH levels were clearly elevated in these animals (Fig. 6). Interestingly, no difference was found in the number of GnRH mRNA-expressing cells in the OVX + E2 + P 4 group in spite of high LH levels.
Discussion In situ hybridization histochemistry is a particularly appropriate technique for examining GnRH gene expres-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
100-
SI o i
50-
10-
5-I
Saline
Pentobarbital
FIG. 5. Effects of pentobarbital on GnRH gene expression and serum LH levels (n = 4 for each group). Proestrous rats were treated with either pentobarbital or saline at 1100 h and killed at 1800 h. The number of GnRH mRNA-expressing cells (GN/S) is expressed as a percentage of GN/S in saline-treated animals. Asterisks indicate P < 0.01 compared to saline-treated controls. sion. U n l i k e m a n y o t h e r h y p o t h a l a m i c
neuropeptides
that are tightly localized to specific areas, GnRH-expressing cells are scattered throughout much of the hypothalamus and express GnRH mRNA at very low levels. Therefore, the ability to individually examine these neurons for GnRH mRNA provides important information about the modulation of GnRH gene expression at the cellular level. We used an [35S]UTP-labeled RNA probe complementary to the GnRH mRNA to localize GnRH mRNA-expressing cells in the rat brain. We found a distribution of GnRH cells very similar to that reported using GnRH immunocytochemistry. Indeed, it has recently been possible to colocalize GnRH mRNA and protein to the same neuronal cells (23). This study reported a close relationship between the numbers of GnRH-immunoreactive cells and GnRH mRNA-expressing cells in female rats, suggesting that most GnRH cells were visualized by GnRH in situ hybridization. In the present study we decided to use the number of GnRH mRNA-expressing cells rather than the intensity of hybridization to individual cells as an index of GnRH gene expression for several reasons. Firstly, we found that the intensity of hybridization varied among individual cells located in close proximity. Similar observations have been reported by others (15, 20, 23). Secondly, in
369
cells expressing high levels of GnRH mRNA, silver grains extensively overlapped, making grain counting or density measurement very difficult. Finally, the method of counting cells after in situ hybridization has been successfully used by others to measure changes in gene expression for neuropeptides such as GnRH (19, 24) and vasopressin (25). In cycling rats we observed that the number of GnRH mRNA-expressing cells remained fairly constant from estrus to diestrus, but did change significantly during proestrus. In particular, there was a pronounced increase in the number of hybridizing cells at the time of the preovulatory LH surge. Preovulatory GnRH hypersecretion is thought to be preceded by an increase in the synthesis and/or accumulation of GnRH peptide in GnRH neurons (5, 26), which is likely to reflect the rate of prepro-GnRH synthesis in the GnRH cell body (20). These changes may be accounted for by the changes in GnRH gene expression that we have observed during this critical period. The gradual decrease (statistically not significant) in GnRH gene expression observed before the LH surge may result from feedback inhibition on GnRH synthesis due to the large accumulation of GnRH peptide in neurons at this time. The increase (statistically significant) in GnRH gene expression observed after the onset of the LH surge might, in turn, result from GnRH depletion, providing a compensatory mechanism for restoring GnRH to basal levels. Increased secretory activity has been shown to increase the synthesis of other neuropeptides in hypothalamic magnocellular neurons (27). Another recent study (28) also finds changes in GnRH gene expression during proestrus, in agreement with the present data. These investigators demonstrated that the intensity of hybridization to single neurons at the level of the OVLT changes during the cycle, decreasing throughout diestrus and early proestrus, then increasing on the evening of proestrus coincident with the LH surge. In this study, however, GnRH mRNA levels, as determined by grain intensity over individual cells, were no higher on proestrous evening than at other times during the cycle. In contrast, we found a significant elevation in the number of GnRH mRNA-expressing cells on proestrous evening. These somewhat different results might be due to the assays used to measure GnRH gene expression in the two studies, as described above. In addition, the hypothalamic areas analyzed were slightly different; we have examined the complete POA-OVLT-DBB area. Although the area at the level of the OVLT contains many GnRH cells, changes in gene expression in these GnRH cells may not solely account for the mechanism involved in the generation of the preovulatory LH surge, since GnRH cells in the POA region are likely to be involved in this event, as discussed earlier (2, 3, 5, 8-10).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
370
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE 100-
Endo • 1990 Vol 127«Nol
1100 h 1800 h
o o
oo CM
LJ
50-
FIG. 6. Effect of ovarian steroids on GnRH gene expression and serum LH levels in OVX animals killed at two different times, 1100 and 1800 h (n = 4 for each group). The number of GnRH mRNA-expressing cells (GN/S) is expressed as a percentage of GN/S at 1800 h in OVX + E2 rats. Asterisks indicate P < 0.05 compared to the other groups.
15-
c
10-
5-
OVX
Moreover, many, but not all, female rats with a radiofrequency-induced lesion in the OVLT showed acyclicity. The preovulatory LH surge in rats showing cyclicity after the surgery was reduced, but not completely blocked (29), suggesting that regions other than the OVLT were involved in generation of the LH surge. Lastly, immunocytochemical quantification of GnRH cells indicates that GnRH content changes in the POA and median eminence, but not in the OVLT, during the estrogen- plus progesterone-induced LH surge in the OVX animal (30). Pentobarbital, a barbiturate, is an anticonvulsant that decreases neuronal activity in the central nervous system. Our data demonstrating a blockade of the preovulatory LH surge by this barbiturate confirm previous studies showing that barbiturate administration during the early afternoon of proestrus blocks the preovulatory LH surge (26, 31, 32) and subsequently blocks ovulation (31). The effect of barbiturates on the LH surge has been thought to be due to a block of GnRH secretion from the hypothalamus. Indeed, GnRH levels in portal blood of barbiturate-treated animals are lower than those in normal rats (33). Our data demonstrating that pentobarbital administration at 1100 h on proestrus blocks the increase in GnRH gene expression normally observed at the time of the LH surge indicate that GnRH hypersecretion is an important factor in stimulating GnRH gene expres-
OVX+Er
OVX+E2+P4
sion. It remains possible that pentobarbital interrupts neuronal transmission and generally affects the synthetic activity of the GnRH neurons (34). It is widely accepted that ovarian steroids are important for modulating GnRH release from the hypothalamus. The role of estrogen in inducing the LH surge (35) has been assumed to be to increase GnRH secretion (5, 36). Because estrogen acts to modulate the transcription of specific genes, it might directly induce GnRH gene expression. Several studies have examined the effects of gonadal steroids on GnRH gene expression in different experimental conditions (13-20). Our data indicate that in the OVX animal, there is a small stimulatory effect of estrogen on GnRH gene expression. However, it is likely that this represents an indirect effect, since GnRH gene expression was increased in the OVX + E2 group only at the time of the normal LH surge. Indeed, serum LH values were elevated in these OVX + E2 animals. This observation is consistent with the hypothesis that hypothalamic GnRH release preceding the LH surge is closely coupled to the increase in the number of GnRH mRNA-expressing cells. It is possible that the LH surge transmits a short-loop feedback signal for GnRH gene expression to the hypothalamus. In any event, it is unlikely that estrogen has a direct action on GnRH gene expression, since GnRH cells appear not to contain es-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
trogen receptors (37). It is curious that we did not observe a similar increase in the number of GnRH mRNA-expressing cells in OVX + E2 + P 4 animals, in spite of high serum LH levels. Progesterone has been reported to advance the GnRH hypersecretion observed in this animal model (38), and therefore, it is possible that we measured GnRH mRNA at the time when mRNA levels had already returned to basal levels. Alternatively, progesterone may be involved in turning off the LH surge, as described previously (5, 39) and may exert direct negative effects at the level of the hypothalamus, perhaps on GnRH gene expression. Previous studies have found that progesterone can either increase (40) or decrease (18) GnRH gene expression. In summary, our results are most consistent with a model in which GnRH hypersecretion at the time of the preovulatory LH surge results in increased GnRH gene expression late on proestrus. By affecting GnRH secretion, both pentobarbital and ovarian steroids may modulate GnRH gene expression in an indirect fashion. Further analysis at the molecular level will be required to define the signals necessary for regulated expression of the GnRH gene in hypothalamic neurons.
11.
12. 13. 14.
15. 16.
17.
18.
19.
Acknowledgments We wish to thank Dr. A. F. Parlow and the NIDDK for the LH RIA kit, Mr. R. Valadka for performing the LH RIA, and Drs. J. E. Levine and N. B. Schwartz for comments on the manuscript.
20.
21. References 1. Swanzel-Fukuda M, Pfaff DW 1989 Origin of luteinizing hormonereleasing hormone neurons. Nature 338:161 2. Barry J, Hoffman GE, Wray S 1985 LHRH-containing systems. In: Bjorklund AS, Hokefelt T (eds) Handbook of Chemical Neuroanatomy. Elsevier, Amsterdam, vol 4:225 3. Merchenthaler I, Setalo G, Csontos C, Petrusz P, Flerko B, NegroVilar A 1988 Combined retrograde tracing and immunocytochemical identification of luteinizing hormone-releasing hormone and somatostatin-containing neurons projecting to the median eminence of the rat. Endocrinology 125:2812 4. Clayton RN 1989 Gonadotrophin-releasing hormone: its actions and receptors. J Endocrinol 120:11 5. Kalra SP 1986 Neural circuitry involved in the control of LHRH secretion: a model for the preovulatory LH release. In: Ganong WF, Marttini L (eds) Frontiers in Neuroendocrinology. Raven Press, New York, vol 9:31 6. Andrews WV, Maurer R, Conn PM 1988 Stimulation of rat luteinizing hormone /3-messenger RNA levels by gonadotropin-releasing hormone. J Biol Chem 263:13755 7. Wierman ME, Rivier JE, Wang C 1989 Gonadotropin-releasing hormone-dependent regulation of gonadotropin subunit messenger ribonucleic acid levels in the rat. Endocrinology 124:272 8. Phelps CP, Krieg RT, Sawyer CH 1976 Spontaneous and electrochemically stimulated changes in plasma LH in the female rat following hypothalamic deafferentation. Brain Res 101:239 9. Kalra SP 1976 Tissue levels of luteinizing hormone-releasing hormone in the preoptic area and hypothalamus and serum concentrations of gonadotropins following anterior hypothalamic deafferentation and estrogen treatment of the female rat. Endocrinology 99:101 10. Palkovits M 1979 Effect of surgical deafferentation on the neuro-
22.
23.
24.
25. 26.
27. 28.
29.
371
transmitter and hormone content of the hypothalamus. Neuroendocrinology 29:140 Adelman JP, Mason AJ, Hayflick JS, Seeburg PH 1986 Isolation of the gene and hypothalamic cDNA for the common precursor of gonadotropin-releasing hormone and prolactin release-inhibiting factor in human and rat. Proc Natl Acad Sci USA 83:179 Bond CT, Hayflick JS, Seeburg PH, Adelman JP 1989 The rat gonadotropin-releasing hormone: SH locus: structure and hypothalamic expression. Mol Endocrinol 3:1257 Pfaff DW 1986 Gene expression in hypothalamic neurons: luteinizing hormone-releasing hormone. J Neurosci Res 16:109 Roberts JL, Dutlow CM, Jakubowski M, Blum M, Millar RP 1989 Estradiol stimulates preoptic area-anterior hypothalamic proGnRH-GAP gene expression in ovariectomized rats. Mol Brain Res 6:127 Rothfeld J, Hejtmancik JF, Conn PM, Pfaff DW 1989 In situ hybridization for LHRH mRNA following estrogen treatment. Mol Brain Res 6:121 Wray S, Zoeller RT, Gainer H 1989 Differential effects of estrogen on luteinizing hormone-releasing hormone gene expression in slice explant cultures prepared from specific rat forebrain regions. Mol Endocrinol 3:1197 Zoeller RT, Seeburg PH, Young III WS 1988 In situ hybridization histochemistry for messenger ribonucleic acid (mRNA) encoding gonadotropin-releasing hormone (GnRH): effect of estrogen on cellular levels of GnRH mRNA in female rat brain. Endocrinology 122:2570 Toranzo D, Dupont E, Simard J, Labrie C, Couet J, Labrie F, Pelletier G 1989 Regulation of pro-gonadotropin-releasing hormone gene expression by sex steroids in the brain of male and female rats. Mol Endocrinol 3:1748 Malik KF, Silverman AJ, Morrel JI, LHRH mRNA: neuronal distribution and steroid hormone regulation. 19th Annual Meeting of the Society for Neuroscience, Pheonix AZ, 1989, p 189 (Abstract) Kelly MJ, Garrett J, Bosch MA, Rosellu CE, Douglass J, Adelman JP, Ronnekleiv OK 1989 Effects of ovariectomy on GnRH mRNA, proGnRH, and GnRH levels in the preoptic hypothalamus of the female rat. Neuroendocrinology 49:88 Woodruff TK, Meunier H, Jones PB, Hsueh JW, Mayo KE 1987 Rat inhibin: molecular cloning of a- and /S-subunit complementary deoxyribonucleic acids and expression in the ovary. Mol Endocrinol 1:561 Melton DA, Krieg PA, Rebagliata MR, Maniatis T, Zinn K, Green MR 1984 Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucleic Acids Res 12:7035 Ronnekleiv OK, Naylor BR, Bond CT, Adelman JP 1989 Combined immunohistochemistry for gonadotropin-releasing hormone (GnRH) and pro-GnRH and ire situ hybridization for GnRH mRNA in the rat brain. Mol Endocrinol 3:363 Wiemann JN, Steiner RA, Recruitment of neurons expressing gonadotropin-releasing hormone messenger RNA following castration in the adult male rat. 71st Annual Meeting of The Endocrine Society, Seattle WA, 1989, p 255 Miller MA, Urban JH, Dorsa DM 1989 Steroid dependency of vasopressin neurons in the bed nucleus of the stria terminalis by ire situ hybridization. Endocrinology 125:2335 Wise PM, Ranee N, Selmanoff M, Barraclough CA 1981 Changes in radioimmunoassayable luteinizing hormone-releasing hormone in discrete brain areas of the rat at various times on proestrus, diestrous day 1, and after phenobarbital administration. Endocrinology 108:2179 Lightman SL, Young III WS 1987 Changes in hypothalamic proenkephalin A mRNA following stress and opiate withdrawal. Nature 328:643 Zoeller RT, Young III WS 1988 Changes in cellular levels of messenger ribonucleic acid encoding gonadotropin-releasing hormone in the anterior hypothalamus of female rats during the estrous cycle. Endocrinology 123:1688 Piva F, Limonta P, Martini L1982 Role of the organum vasculosum laminae terminalis in the control of gonadotropin secretion in rats. J Endocrinol 93:344
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
372
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
30. Rothfeld JM, Gross D 1985 Gonadotropin-releasing hormone within the organum vasculosum of the lamina terminalis in the ovariectomized, estrogen/progesterone treated rat: a quantitative immunocytochemical study using image analysis. Brain Res 338:L309 31. Everett JW, Sawyer CH 1950 24 hour periodicity in the "LHrelease apparatus" of female rats, disclosed by barbiturate sedatives. Endocrinology 47:198 32. Legan SJ, Karsch FJ 1975 A daily signal for the LH surge in the rat. Endocrinology 96:57 33. Eskay RL, Mical RS, Porter JC 1977 Relationship between luteinizing hormone-releasing hormone concentration in hypophyseal portal blood and luteinizing hormone release in intact, castrated and electrochemically stimulated rats. Endocrinology 100:263 34. Ranee N, Barraclough CA 1981 Effects of phenobarbital on hypothalamic LHRH and catecholamine turnover rates in proestrous rats. Proc Soc Exp Biol Med 166:425
Endo • 1990 Vol 127 • No 1
35. Legan SJ, Coon GA, Karsch FJ 1975 Role of estrogen as initiator of daily LH surges in the ovariectomized rat. Endocrinology 96:50 36. Dluzen DE, Ramirez VD 1986 In vivo LHRH output of ovariectomized rats following estrogen treatment. Neuroendocrinology 118:1110 37. Shivers BD, Harlan RE, Morrel JI, Pfaff DW 1983 Absence of oestradiol concentration in cell nuclei of LHRH immunoreactive neurons. Nature 304:345 38. McPherson JC, Costoff A, Mahesh VB 1975 Influence of estrogenprogesterone combination on gonadotropin secretion in castrated female rats. Endocrinology 91:771 39. Smith MS, Fox SR, Chatterton RT 1989 Role of proestrous progesterone secretion in suppressing basal pulsatile LH secretion during estrus of the estrous cycle. Neuroendocrinology 50:308 40. Kim K, Lee BJ, Park Y, Cho W 1989 Progesterone increases messenger ribonucleic acid (mRNA) encoding luteinizing hormonereleasing hormone (LHRH) level in the hypothalamus of ovariectomized estradiol-primed prepubertal rats. Mol Brain Res 6:151
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
Vol. 127, No. 1 Printed in U.S.A.
Gonadotropin-Releasing Hormone Gene Expression during the Rat Estrous Cycle: Effects of Pentobarbital and Ovarian Steroids* OK-KYONG PARK, SAJIV GUGNEJA, AND KELLY E. MAYO Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, Euanston, Illinois 60208
ABSTRACT. Although hypothalamic GnRH release is known to be modulated by neural and hormonal factors, the relationship between altered GnRH secretion and GnRH synthesis remains unclear. In an attempt to address this question, we examined GnRH gene expression in the rat hypothalamus using in situ hybridization histochemistry. An 25S-labeled antisense RNA probe was used to identify neurons expressing GnRH mRNA in an area that included the diagonal band of Broca, the organum vasculosum of the lamina terminalis, and the preoptic area. The number of GnRH mRNA-expressing cells was determined at various times during the rat estrous cycle. During proestrus, the number of GnRH mRNA-expressing cells decreased somewhat at 1400-1600 h, increased significantly at 1800 h (the time of the LH surge), then gradually returned to basal levels at 2200 h. Expression did not change substantially at other times during the estrous cycle. To understand this close temporal relationship between the LH surge and increased GnRH mRNA levels, we examined GnRH gene expression in proestrous animals in which
the LH surge was blocked with pentobarbital. Pentobarbital treatment blocked the increase in the number of GnRH mRNAexpressing cells normally observed at 1800 h in saline-treated controls, suggesting that the increase in GnRH gene expression is closely coupled to secretion of GnRH from the hypothalamus. Finally, we addressed the question of whether ovarian steroids have direct effects on GnRH gene expression by examining GnRH mRNA levels in ovariectomized steroid-treated rats at a time before (1100 h) and a time after (1800 h) hypothalamic GnRH hypersecretion. At 1100 h, no significant changes were observed, but at 1800 h, estrogen-treated rats showed a significant increase in both the number of GnRH mRNA-expressing cells and serum LH levels. This suggests that estrogen influences GnRH gene expression indirectly, perhaps by altering hypothalamic GnRH release. Our results in each of these models suggest that GnRH mRNA levels increase in response to GnRH hypersecretion at the time of the LH surge. (Endocrinology 127: 365372,1990)
G
in profound alterations in the estrous cycle (8). It also results in topological changes in GnRH content, such as elevation in the area anterior to the optic chiasm (9) and severe depletion in the median eminence (10). After the isolation of GnRH cDNA clones and the prediction of the GnRH precursor sequence (11, 12), several groups have used in situ hybridization, Northern or dot blot RNA analysis, and solution hybridization to determine changes in GnRH gene expression in animals treated with different hormones. These results, however, have been difficult to interpret. For example, estrogen, which induces the LH surge, has been reported to exert stimulatory (13-15), inhibitory (16-18), or no (19) effects on GnRH gene expression. In addition, elimination of gonadal steroid feedback by ovariectomy (OVX) is reported to have diverse effects on GnRH gene expression (14, 18, 20). These differing results might be attributable to differences in the sensitivity of techniques, anatomical regions analyzed, and/or time points used in these studies. To understand dynamic changes in GnRH gene expression in more detail, we attempted to examine GnRH gene
nRH NEURONS originate from the medial offactory placode during embryonic development (1) and migrate predominantly into the diagonal band of Broca (DBB), the organum vasculosum of the lamina terminalis (OVLT), and the preoptic area (POA) (2). GnRH cells in the OVLT-POA-DBB region are believed to be directly involved in controlling the function of pituitary gonadotropes, since a majority of GnRH cells from these areas send projections to the median eminence (3) and secrete GnRH into the hypophysial portal bloodstream. GnRH activates pituitary GnRH receptors (4) and thereby modulates the synthesis and/or release of gonadotropins (5-7). GnRH cells in the OVLT-POA area are clearly important for normal reproductive function, since deafferentation of the hypothalamus results Received February 14, 1990. Address all correspondence and requests for reprints to: Dr. Kelly E. Mayo, Department of Biochemistry, Molecular Biology, and Cell Biology, Northwestern University, 2153 Sheridan Road, Evanston, Illinois 60208. *This work was supported by a development award from the McKnight Fund for Neuroscience and NIH Grants NS-24439 and HD21921.
365
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
366
expression at many time points in intact rats during the estrous cycle, in proestrous rats treated with pentobarbital, and in OVX animals treated with combinations of ovarian steroids. A sensitive and specific in situ hybridization assay was used to determine the number of GnRH mRNA-expressing neurons in the POA-OVLT-DBB area, while serum LH levels were used as an index of hypothalamic GnRH release. Materials and Methods Animals Intact Sprague-Dawley female rats (200-240 g BW) were purchased from Charles River Breeding Laboratories (Wilmington, MA) and kept in a photoperiod of 14 h of light and 10 h of darkness, with lights on at 0500 h. Food and water were available ad libitum. Stages of the estrous cycle of a group of rats were monitored by daily vaginal lavage, and only those rats demonstrating at least two consecutive 4-day cycles were used in this study. Another group of rats was bilaterally OVX under methoxyflurane anesthesia. Three different experimental groups were included in this study: intact rats during various stages of the estrous cycle, pentobarbital-treated proestrous animals, and steroid-replaced OVX animals. Estrous cycle experiment. Rats (n = 4 or 5 for each time point) were killed at different time intervals (at 1100 and 1800 h on estrus, metestrus, and diestrus and at 0900, 1100, 1400, 1600, 1800, 2000, and 2200 h on proestrus). After decapitation, trunk blood was collected for subsequent serum LH determination, and the brain was rapidly removed, frozen on dry ice, and stored at -70 C until used. Pentobarbital experiment. At 1100 h on proestrus, rats (n = 4 in each group) were injected ip with either sodium pentobarbital (100 mg/kg BW) or saline. At 1800 h, rats were decapitated, and brains and trunk blood were collected as described above. Steroid-replacement experiment. Three groups of rats were used for this experiment: OVX, OVX estrogen-primed (OVX + E2), and OVX estrogen- plus progesterone-primed (OVX + E2 + P4) rats. Each of these groups was examined at the time points of 1100 and 1800 h (n = 4 in each group). At 0930-1000 h 7-9 days after OVX surgery, animals were implanted with either a crystalline E2 capsule (5 mm; Dow-Corning Silastic tubing, Medfield, MA; id, 0.062 in.; od, 0.095 in.) or an empty capsule. At 0900 h 2 days later, animals were sc injected with either progesterone (1.5 mg in mineral oil) or vehicle only. Rats were decapitated at either 1100 or 1800 h, and brains and trunk blood were collected as described above.
NaCl, 30 mM sodium citrate, pH 7.0) for 5 min, dipped in distilled deionized H2O and 0.1 M triethanolamine, and incubated in 0.25% acetic anhydride in 0.1 M triethanolamine for 10 min. Sections were dehydrated through an ethanol series and vacuum dried until hybridization. The GnRH clone was isolated from a rat hypothalamic cDNA library by screening with an oligonucleotide corresponding to the mature GnRH decapeptide. The largest clone isolated contained a 575-basepair (bp) insert that includes 128 bp of 5' nontranslated sequences, the complete coding region for the 92-amino acid GnRH precursor, 146 bp of 3' nontranslated sequennes, and a 25-bp poly (A) tract. The sequence of the clone corresponds to that previously reported by others (11, 12) for the rat GnRH cDNA and gene. The cDNA insert was subcloned in both orientations into pGEM-4 for probe preparation (see Fig. 1). Antisense and sense [35S]UTP-labeled RNA probes were synthesized using SP6 polymerase (22) from GnRH-C digested with HindI.II (Fig. 1A) and GnRH-A digested with Kpnl (Fig. IB), respectively. Probe (107 cpm/ml hybridization buffer) was applied to the slides, and they were coverslipped and hybridized at 47 C for 15 h. After hybridization, the coverslips were removed in 4 x SSC, and sections were treated with RNase-A (20 ng/va\) at 37 C for 30 min, washed to a stringency of 0.25 x SSC at 55 C, and dehydrated through an ethanol series. Sections were exposed on Kodak XAR-5 film (Eastman Kodak, Rochester, NY) overnight, and then dipped in Kodak NTB-2 liquid emulsion and exposed for 2 weeks. After developing and fixing, sections were stained with cresyl violet, observed, and photographed using a Nikon (Garden City, NY) optiphot microscope. The microscope was equipped with a Javelin (Torrance, CA) solid state camera and Sony (Park Ridge, NJ) monitor to facilitate cell counting.
LHRIA LH concentrations were determined using an ovine:rat RIA system employing a NIDDK kit (NIH LH S-25 as standard
T7
100bp
Antisense RNA Probe
b
GnRH in situ hybridization Coronal sections (20 j^m) of the brain throughout the DBB, OVLT, and POA were cut on a Reichert 820 cryostat (Reichert, Buffalo, NY) at -20 C and mounted onto gelatin and poly-Llysine-coated glass slides for in situ hybridization, as described previously (21). Briefly, brain sections were fixed in 5% paraformaldehyde (pH 7.5) for 5 min, washed in 2 x SSC (300 mM
Endo • 1990 Vol 127 • No 1
t> * if
1— PGEM4 100 bp
Sense RNA Probe FIG. 1. Schematic structure of the rat GnRH cDNA clone used in this study. A full-length GnRH cDNA was subcloned into the vector pGEM4 either in an antisense (A) or a sense (B) orientation.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE and antirat LH antibody S-10). Results are expressed as nanograms per milliliter NIH LH S-25 standard. The intra- and interassay coefficients of variation were 5.0% and 6.0%, respectively.
100X:OVLT/POA
367 500X: OVLT/POA
Data analysis GnRH mRNA-expressing cells were manually counted in 44-46 sections (20 /*m) from an area encompassing the DBB, OVLT, and POA. The average number of GnRH mRNAexpressing cells per brain section in the POA-OVLT-DBB area (GN/S) was determined for each animal. The number of GnRH mRNA-expressing cells in cycling animals from five independent experiments was expressed as a percentage of the value at 1800 h on metestrus in each experiment. The number of hybridizing cells per section at 1800 h on metestrus ranged between 19.9-22.4 in the five experiments (mean, 20.9 ± 0.8). For the pentobarbital and steroid replacement experiments, the values at 1800 h in saline-treated and OVX + E2 rats were considered as 100%. GN/S ranged between 22.9-33.3 at 1800 h on proestrus in saline-treated animals (mean, 28.2 ± 3.2) and between 23.0-28.9 at 1800 h in OVX + E2 animals (mean, 25.4 ± 1.2). One-way analysis of variance followed by the post-hoc test were used for statistical analysis of data from the cycling and steroid replacement experiments, while the group Student t test was used for data from the pentobarbital experiment. P < 0.05 was considered significant.
Results Figure 1 illustrates the structure of a full-length GnRH cDNA clone that was isolated from a rat hypothalamic cDNA library and subcloned into the plasmid vector pGEM-4 in either an antisense (Fig. 1A) or a sense (Fig. IB) orientation. This clone was used to synthesize riboprobes for in situ hybridization histochemistry in the rat brain. Examination of brain sections with an [35S]UTPlabeled antisense RNA probe revealed hybridizing cells in the area of the DBB, OVLT, and POA, areas known as loci for GnRH-immunoreactive cells. The hybridizing cells had a random morphology typical of GnRH cells: oval, fusiform, and spindly shaped. Figure 2 shows an example of brain sections containing GnRH mRNAexpressing cells. Silver grains indicative of positive hybridization appear to be deposited over individual cells (Fig. 2, A and C) and are easily discernable over background, particularly using darkfield microscopy (Fig. 2, B and D). The specificity of hybridization was tested in sequential brain sections that were taken at the level of the OVLT-POA and hybridized with either an antisense or a sense strand GnRH riboprobe. Figure 3A shows several dozen cells expressing GnRH mRNA observed using the antisense probe; no hybridization was observed using the sense probe (Fig. 3B). In addition, the GnRH antisense probe did not hybridize to a variety of control tissues examined, including liver and ovary (data not shown).
FIG. 2. Examples of GnRH cells in the POA after hybridization to a GnRH antisense RNA probe. A and B are photographs taken at X100 magnification; C and D are photographs taken at X500 magnification. The upper panels were photographed using brightfield optics, and the lower panels used darkfield optics.
40X: OVLT/POA
o C Q) CO
o ol a> (/> c a> (A
FIG. 3. Darkfield microscopy of adjacent brain sections at the level of the OVLT-POA. Sections were hybridized with either an [35S]UTPlabeled antisense RNA or an [35S]UTP-labeled sense RNA probe. Photographs were taken at X40 magnification.
To determine whether GnRH gene expression changes during the estrous cycle, we examined GnRH mRNA levels in animals at different times throughout the 4-day cycle. As an index of GnRH mRNA expression, we counted the number of hybridizing cells in an area that included the DBB, OVLT, and POA. We also measured serum LH levels in the animals used for this study. These
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
368
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
Endo • 1990 Vol 127 • No 1
200-1
o o CO CO
•8 1 0 ° H FIG. 4. Changes in the number of GnRH mRNA-expressing cells and in serum LH levels during the rat estrous cycle (n = 4 or 5 for each group). A, GN/S is the average number of GnRH mRNA-positive cells per brain section; values are expressed as a percentage of GN/S at 1800 h on metestrus. B shows serum LH values in the corresponding animals. Asterisks indicate P < 0.05 compared to the other time points.
z o
B c 10-
1200
1200
Metestrus
data are presented in Fig. 4. Figure 4A shows that the number of GnRH mRNA-expressing cells, expressed as a percentage of the value at 1800 h on metestrus, is maintained at a fairly constant level throughout the estrous cycle, except during proestrus. During proestrus, the number of GnRH mRNA-expressing cells decreases somewhat at 1400-1600 h, then increases about 2-fold (P < 0.01) at 1800 h, and gradually returns to basal levels at 2200 h. Serum LH levels in these animals were significantly elevated by 1600 h on proestrus, 2 h before the rise in the number of GnRH mRNA-expressing cells, and remained high throughout the evening of proestrus (Fig. 4B). The close temporal correlation between the increased number of GnRH mRNA-expressing cells and the onset of the preovulatory LH surge prompted us to further investigate the relationship between these events. We used pentobarbital to eliminate the LH surge in proestrous rats and examined GnRH mRNA levels in these animals. Figure 5 shows the effects of pentobarbital on serum LH levels and GnRH gene expression. As expected, pentobarbital treatment completely blocked the preovulatory LH surge (P < 0.01), which remained unaffected in saline-treated animals. The number of GnRH mRNA-expressing cells was also significantly lower (P
1200
Diestrus Proestrus Rat Colony Time (h)
1200
Estrus
< 0.01) in pentobarbital-treated rats compared to that in saline-treated animals (65.1 ± 7.6% us. 100.0 ± 4.3%; Fig. 5). Finally, we examined the effect of ovarian steroid feedback on GnRH gene expression. Animals were OVX for 7-9 days, then replaced with either estrogen alone or estrogen plus progesterone. Rats were subsequently examined at two times of day: 1100 h, a time when the number of GnRH mRNA-expressing cells has not significantly changed in proestrous animals, and 1800 h, a time when the number of GnRH-mRNA-expressing cells is maximal in proestrous animals (Fig. 4). As shown in Fig. 6, no differences were seen among the three groups at 1100 h. However, a small but significant (P < 0.05) increase in the number of GnRH mRNA-expressing cells was observed in the OVX + E2 group. Serum LH levels were clearly elevated in these animals (Fig. 6). Interestingly, no difference was found in the number of GnRH mRNA-expressing cells in the OVX + E2 + P 4 group in spite of high LH levels.
Discussion In situ hybridization histochemistry is a particularly appropriate technique for examining GnRH gene expres-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
100-
SI o i
50-
10-
5-I
Saline
Pentobarbital
FIG. 5. Effects of pentobarbital on GnRH gene expression and serum LH levels (n = 4 for each group). Proestrous rats were treated with either pentobarbital or saline at 1100 h and killed at 1800 h. The number of GnRH mRNA-expressing cells (GN/S) is expressed as a percentage of GN/S in saline-treated animals. Asterisks indicate P < 0.01 compared to saline-treated controls. sion. U n l i k e m a n y o t h e r h y p o t h a l a m i c
neuropeptides
that are tightly localized to specific areas, GnRH-expressing cells are scattered throughout much of the hypothalamus and express GnRH mRNA at very low levels. Therefore, the ability to individually examine these neurons for GnRH mRNA provides important information about the modulation of GnRH gene expression at the cellular level. We used an [35S]UTP-labeled RNA probe complementary to the GnRH mRNA to localize GnRH mRNA-expressing cells in the rat brain. We found a distribution of GnRH cells very similar to that reported using GnRH immunocytochemistry. Indeed, it has recently been possible to colocalize GnRH mRNA and protein to the same neuronal cells (23). This study reported a close relationship between the numbers of GnRH-immunoreactive cells and GnRH mRNA-expressing cells in female rats, suggesting that most GnRH cells were visualized by GnRH in situ hybridization. In the present study we decided to use the number of GnRH mRNA-expressing cells rather than the intensity of hybridization to individual cells as an index of GnRH gene expression for several reasons. Firstly, we found that the intensity of hybridization varied among individual cells located in close proximity. Similar observations have been reported by others (15, 20, 23). Secondly, in
369
cells expressing high levels of GnRH mRNA, silver grains extensively overlapped, making grain counting or density measurement very difficult. Finally, the method of counting cells after in situ hybridization has been successfully used by others to measure changes in gene expression for neuropeptides such as GnRH (19, 24) and vasopressin (25). In cycling rats we observed that the number of GnRH mRNA-expressing cells remained fairly constant from estrus to diestrus, but did change significantly during proestrus. In particular, there was a pronounced increase in the number of hybridizing cells at the time of the preovulatory LH surge. Preovulatory GnRH hypersecretion is thought to be preceded by an increase in the synthesis and/or accumulation of GnRH peptide in GnRH neurons (5, 26), which is likely to reflect the rate of prepro-GnRH synthesis in the GnRH cell body (20). These changes may be accounted for by the changes in GnRH gene expression that we have observed during this critical period. The gradual decrease (statistically not significant) in GnRH gene expression observed before the LH surge may result from feedback inhibition on GnRH synthesis due to the large accumulation of GnRH peptide in neurons at this time. The increase (statistically significant) in GnRH gene expression observed after the onset of the LH surge might, in turn, result from GnRH depletion, providing a compensatory mechanism for restoring GnRH to basal levels. Increased secretory activity has been shown to increase the synthesis of other neuropeptides in hypothalamic magnocellular neurons (27). Another recent study (28) also finds changes in GnRH gene expression during proestrus, in agreement with the present data. These investigators demonstrated that the intensity of hybridization to single neurons at the level of the OVLT changes during the cycle, decreasing throughout diestrus and early proestrus, then increasing on the evening of proestrus coincident with the LH surge. In this study, however, GnRH mRNA levels, as determined by grain intensity over individual cells, were no higher on proestrous evening than at other times during the cycle. In contrast, we found a significant elevation in the number of GnRH mRNA-expressing cells on proestrous evening. These somewhat different results might be due to the assays used to measure GnRH gene expression in the two studies, as described above. In addition, the hypothalamic areas analyzed were slightly different; we have examined the complete POA-OVLT-DBB area. Although the area at the level of the OVLT contains many GnRH cells, changes in gene expression in these GnRH cells may not solely account for the mechanism involved in the generation of the preovulatory LH surge, since GnRH cells in the POA region are likely to be involved in this event, as discussed earlier (2, 3, 5, 8-10).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
370
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE 100-
Endo • 1990 Vol 127«Nol
1100 h 1800 h
o o
oo CM
LJ
50-
FIG. 6. Effect of ovarian steroids on GnRH gene expression and serum LH levels in OVX animals killed at two different times, 1100 and 1800 h (n = 4 for each group). The number of GnRH mRNA-expressing cells (GN/S) is expressed as a percentage of GN/S at 1800 h in OVX + E2 rats. Asterisks indicate P < 0.05 compared to the other groups.
15-
c
10-
5-
OVX
Moreover, many, but not all, female rats with a radiofrequency-induced lesion in the OVLT showed acyclicity. The preovulatory LH surge in rats showing cyclicity after the surgery was reduced, but not completely blocked (29), suggesting that regions other than the OVLT were involved in generation of the LH surge. Lastly, immunocytochemical quantification of GnRH cells indicates that GnRH content changes in the POA and median eminence, but not in the OVLT, during the estrogen- plus progesterone-induced LH surge in the OVX animal (30). Pentobarbital, a barbiturate, is an anticonvulsant that decreases neuronal activity in the central nervous system. Our data demonstrating a blockade of the preovulatory LH surge by this barbiturate confirm previous studies showing that barbiturate administration during the early afternoon of proestrus blocks the preovulatory LH surge (26, 31, 32) and subsequently blocks ovulation (31). The effect of barbiturates on the LH surge has been thought to be due to a block of GnRH secretion from the hypothalamus. Indeed, GnRH levels in portal blood of barbiturate-treated animals are lower than those in normal rats (33). Our data demonstrating that pentobarbital administration at 1100 h on proestrus blocks the increase in GnRH gene expression normally observed at the time of the LH surge indicate that GnRH hypersecretion is an important factor in stimulating GnRH gene expres-
OVX+Er
OVX+E2+P4
sion. It remains possible that pentobarbital interrupts neuronal transmission and generally affects the synthetic activity of the GnRH neurons (34). It is widely accepted that ovarian steroids are important for modulating GnRH release from the hypothalamus. The role of estrogen in inducing the LH surge (35) has been assumed to be to increase GnRH secretion (5, 36). Because estrogen acts to modulate the transcription of specific genes, it might directly induce GnRH gene expression. Several studies have examined the effects of gonadal steroids on GnRH gene expression in different experimental conditions (13-20). Our data indicate that in the OVX animal, there is a small stimulatory effect of estrogen on GnRH gene expression. However, it is likely that this represents an indirect effect, since GnRH gene expression was increased in the OVX + E2 group only at the time of the normal LH surge. Indeed, serum LH values were elevated in these OVX + E2 animals. This observation is consistent with the hypothesis that hypothalamic GnRH release preceding the LH surge is closely coupled to the increase in the number of GnRH mRNA-expressing cells. It is possible that the LH surge transmits a short-loop feedback signal for GnRH gene expression to the hypothalamus. In any event, it is unlikely that estrogen has a direct action on GnRH gene expression, since GnRH cells appear not to contain es-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
trogen receptors (37). It is curious that we did not observe a similar increase in the number of GnRH mRNA-expressing cells in OVX + E2 + P 4 animals, in spite of high serum LH levels. Progesterone has been reported to advance the GnRH hypersecretion observed in this animal model (38), and therefore, it is possible that we measured GnRH mRNA at the time when mRNA levels had already returned to basal levels. Alternatively, progesterone may be involved in turning off the LH surge, as described previously (5, 39) and may exert direct negative effects at the level of the hypothalamus, perhaps on GnRH gene expression. Previous studies have found that progesterone can either increase (40) or decrease (18) GnRH gene expression. In summary, our results are most consistent with a model in which GnRH hypersecretion at the time of the preovulatory LH surge results in increased GnRH gene expression late on proestrus. By affecting GnRH secretion, both pentobarbital and ovarian steroids may modulate GnRH gene expression in an indirect fashion. Further analysis at the molecular level will be required to define the signals necessary for regulated expression of the GnRH gene in hypothalamic neurons.
11.
12. 13. 14.
15. 16.
17.
18.
19.
Acknowledgments We wish to thank Dr. A. F. Parlow and the NIDDK for the LH RIA kit, Mr. R. Valadka for performing the LH RIA, and Drs. J. E. Levine and N. B. Schwartz for comments on the manuscript.
20.
21. References 1. Swanzel-Fukuda M, Pfaff DW 1989 Origin of luteinizing hormonereleasing hormone neurons. Nature 338:161 2. Barry J, Hoffman GE, Wray S 1985 LHRH-containing systems. In: Bjorklund AS, Hokefelt T (eds) Handbook of Chemical Neuroanatomy. Elsevier, Amsterdam, vol 4:225 3. Merchenthaler I, Setalo G, Csontos C, Petrusz P, Flerko B, NegroVilar A 1988 Combined retrograde tracing and immunocytochemical identification of luteinizing hormone-releasing hormone and somatostatin-containing neurons projecting to the median eminence of the rat. Endocrinology 125:2812 4. Clayton RN 1989 Gonadotrophin-releasing hormone: its actions and receptors. J Endocrinol 120:11 5. Kalra SP 1986 Neural circuitry involved in the control of LHRH secretion: a model for the preovulatory LH release. In: Ganong WF, Marttini L (eds) Frontiers in Neuroendocrinology. Raven Press, New York, vol 9:31 6. Andrews WV, Maurer R, Conn PM 1988 Stimulation of rat luteinizing hormone /3-messenger RNA levels by gonadotropin-releasing hormone. J Biol Chem 263:13755 7. Wierman ME, Rivier JE, Wang C 1989 Gonadotropin-releasing hormone-dependent regulation of gonadotropin subunit messenger ribonucleic acid levels in the rat. Endocrinology 124:272 8. Phelps CP, Krieg RT, Sawyer CH 1976 Spontaneous and electrochemically stimulated changes in plasma LH in the female rat following hypothalamic deafferentation. Brain Res 101:239 9. Kalra SP 1976 Tissue levels of luteinizing hormone-releasing hormone in the preoptic area and hypothalamus and serum concentrations of gonadotropins following anterior hypothalamic deafferentation and estrogen treatment of the female rat. Endocrinology 99:101 10. Palkovits M 1979 Effect of surgical deafferentation on the neuro-
22.
23.
24.
25. 26.
27. 28.
29.
371
transmitter and hormone content of the hypothalamus. Neuroendocrinology 29:140 Adelman JP, Mason AJ, Hayflick JS, Seeburg PH 1986 Isolation of the gene and hypothalamic cDNA for the common precursor of gonadotropin-releasing hormone and prolactin release-inhibiting factor in human and rat. Proc Natl Acad Sci USA 83:179 Bond CT, Hayflick JS, Seeburg PH, Adelman JP 1989 The rat gonadotropin-releasing hormone: SH locus: structure and hypothalamic expression. Mol Endocrinol 3:1257 Pfaff DW 1986 Gene expression in hypothalamic neurons: luteinizing hormone-releasing hormone. J Neurosci Res 16:109 Roberts JL, Dutlow CM, Jakubowski M, Blum M, Millar RP 1989 Estradiol stimulates preoptic area-anterior hypothalamic proGnRH-GAP gene expression in ovariectomized rats. Mol Brain Res 6:127 Rothfeld J, Hejtmancik JF, Conn PM, Pfaff DW 1989 In situ hybridization for LHRH mRNA following estrogen treatment. Mol Brain Res 6:121 Wray S, Zoeller RT, Gainer H 1989 Differential effects of estrogen on luteinizing hormone-releasing hormone gene expression in slice explant cultures prepared from specific rat forebrain regions. Mol Endocrinol 3:1197 Zoeller RT, Seeburg PH, Young III WS 1988 In situ hybridization histochemistry for messenger ribonucleic acid (mRNA) encoding gonadotropin-releasing hormone (GnRH): effect of estrogen on cellular levels of GnRH mRNA in female rat brain. Endocrinology 122:2570 Toranzo D, Dupont E, Simard J, Labrie C, Couet J, Labrie F, Pelletier G 1989 Regulation of pro-gonadotropin-releasing hormone gene expression by sex steroids in the brain of male and female rats. Mol Endocrinol 3:1748 Malik KF, Silverman AJ, Morrel JI, LHRH mRNA: neuronal distribution and steroid hormone regulation. 19th Annual Meeting of the Society for Neuroscience, Pheonix AZ, 1989, p 189 (Abstract) Kelly MJ, Garrett J, Bosch MA, Rosellu CE, Douglass J, Adelman JP, Ronnekleiv OK 1989 Effects of ovariectomy on GnRH mRNA, proGnRH, and GnRH levels in the preoptic hypothalamus of the female rat. Neuroendocrinology 49:88 Woodruff TK, Meunier H, Jones PB, Hsueh JW, Mayo KE 1987 Rat inhibin: molecular cloning of a- and /S-subunit complementary deoxyribonucleic acids and expression in the ovary. Mol Endocrinol 1:561 Melton DA, Krieg PA, Rebagliata MR, Maniatis T, Zinn K, Green MR 1984 Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucleic Acids Res 12:7035 Ronnekleiv OK, Naylor BR, Bond CT, Adelman JP 1989 Combined immunohistochemistry for gonadotropin-releasing hormone (GnRH) and pro-GnRH and ire situ hybridization for GnRH mRNA in the rat brain. Mol Endocrinol 3:363 Wiemann JN, Steiner RA, Recruitment of neurons expressing gonadotropin-releasing hormone messenger RNA following castration in the adult male rat. 71st Annual Meeting of The Endocrine Society, Seattle WA, 1989, p 255 Miller MA, Urban JH, Dorsa DM 1989 Steroid dependency of vasopressin neurons in the bed nucleus of the stria terminalis by ire situ hybridization. Endocrinology 125:2335 Wise PM, Ranee N, Selmanoff M, Barraclough CA 1981 Changes in radioimmunoassayable luteinizing hormone-releasing hormone in discrete brain areas of the rat at various times on proestrus, diestrous day 1, and after phenobarbital administration. Endocrinology 108:2179 Lightman SL, Young III WS 1987 Changes in hypothalamic proenkephalin A mRNA following stress and opiate withdrawal. Nature 328:643 Zoeller RT, Young III WS 1988 Changes in cellular levels of messenger ribonucleic acid encoding gonadotropin-releasing hormone in the anterior hypothalamus of female rats during the estrous cycle. Endocrinology 123:1688 Piva F, Limonta P, Martini L1982 Role of the organum vasculosum laminae terminalis in the control of gonadotropin secretion in rats. J Endocrinol 93:344
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
372
GnRH GENE EXPRESSION DURING RAT ESTROUS CYCLE
30. Rothfeld JM, Gross D 1985 Gonadotropin-releasing hormone within the organum vasculosum of the lamina terminalis in the ovariectomized, estrogen/progesterone treated rat: a quantitative immunocytochemical study using image analysis. Brain Res 338:L309 31. Everett JW, Sawyer CH 1950 24 hour periodicity in the "LHrelease apparatus" of female rats, disclosed by barbiturate sedatives. Endocrinology 47:198 32. Legan SJ, Karsch FJ 1975 A daily signal for the LH surge in the rat. Endocrinology 96:57 33. Eskay RL, Mical RS, Porter JC 1977 Relationship between luteinizing hormone-releasing hormone concentration in hypophyseal portal blood and luteinizing hormone release in intact, castrated and electrochemically stimulated rats. Endocrinology 100:263 34. Ranee N, Barraclough CA 1981 Effects of phenobarbital on hypothalamic LHRH and catecholamine turnover rates in proestrous rats. Proc Soc Exp Biol Med 166:425
Endo • 1990 Vol 127 • No 1
35. Legan SJ, Coon GA, Karsch FJ 1975 Role of estrogen as initiator of daily LH surges in the ovariectomized rat. Endocrinology 96:50 36. Dluzen DE, Ramirez VD 1986 In vivo LHRH output of ovariectomized rats following estrogen treatment. Neuroendocrinology 118:1110 37. Shivers BD, Harlan RE, Morrel JI, Pfaff DW 1983 Absence of oestradiol concentration in cell nuclei of LHRH immunoreactive neurons. Nature 304:345 38. McPherson JC, Costoff A, Mahesh VB 1975 Influence of estrogenprogesterone combination on gonadotropin secretion in castrated female rats. Endocrinology 91:771 39. Smith MS, Fox SR, Chatterton RT 1989 Role of proestrous progesterone secretion in suppressing basal pulsatile LH secretion during estrus of the estrous cycle. Neuroendocrinology 50:308 40. Kim K, Lee BJ, Park Y, Cho W 1989 Progesterone increases messenger ribonucleic acid (mRNA) encoding luteinizing hormonereleasing hormone (LHRH) level in the hypothalamus of ovariectomized estradiol-primed prepubertal rats. Mol Brain Res 6:151
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 21:03 For personal use only. No other uses without permission. . All rights reserved.
