当前位置: 首页 > 期刊 > 《内分泌学杂志》 > 2005年第9期 > 正文
编号:11295341
Decreased Gonadotropin-Releasing Hormone Neuronal Activity Is Associated with Decreased Fertility and Dysregulation of Food Intake in the Fe
http://www.100md.com 《内分泌学杂志》
     Departments of Physiology (F.G., S.E.l.F., R.I.W., M.F.D.) and Obstetrics, Gynecology and Reproductive Sciences (R.I.W., M.E.M.), University of California, San Francisco, California 94143

    Abstract

    Expression of a cAMP-specific phosphodiesterase in GnRH neurons in the GPR-4 transgenic rat resulted in decreased LH levels and pulse frequency and diminished fertility. We have characterized changes in fertility, adiposity, and reproductive and metabolic hormones with age. Although LH levels were decreased in 3-, 6-, and 9-month-old GPR-4 females relative to wild-type (WT) controls, GPR-4 females did not become anovulatory until 6 months of age. No differences were observed in FSH, estradiol, or androstenedione levels in 3-, 6-, or 9-month-old GPR-4 and WT females. At 9 months of age, GPR-4 females had significantly increased abdominal and sc fat depot weights that were associated with increased leptin and insulin levels not observed in WT females. We tested the hypothesis that metabolic changes observed at 9 months of age were the result of dysregulation of the mechanisms controlling energy balance. Two-month-old female GPR-4 rats placed on a high-energy diet gained weight at a rate significantly greater than WT females and, after 24 d, developed the same metabolic phenotype observed in 9-month-old GRP-4 females (increased abdominal and sc fat associated with elevated leptin and insulin concentrations). Overeating did not correlate with changes in estradiol or androstenedione levels. We conclude that decreased GnRH neuronal activity is closely associated with decreased reproductive function and dysregulation of food intake.

    Introduction

    THE PULSATILE RELEASE of GnRH is the driving force of reproductive function in male and female mammals (1). The establishment of genetically altered transgenic rats with decreased GnRH pulse frequency provided a model for studying the role of pulsatile GnRH release on reproductive function. In a line of transgenic rats (GPR-4) in which expression of a constitutively active cAMP-specific phosphodiesterase, PDE4D1, was genetically targeted to GnRH neurons using the GnRH promoter (2), a large decrease in LH pulse frequency was observed in castrated GPR-4 female and male rats. The rationale for making these animals was based on findings in the GT1-1 GnRH cell line that the intracellular level of cAMP was an important regulator of neuron excitability and the frequency of intrinsic GnRH pulses. This conclusion was based on several observations. Increasing cAMP levels stimulated the secretion of GnRH from cultured GnRH neurons and GT1 GnRH cell lines (3, 4, 5). Increasing cAMP levels in GT1 cells also increased the occurrence of intracellular Ca2+ oscillations (6). Because Ca2+ oscillations are preceded by an action potential, increased Ca2+ oscillations reflect an increase in neuron excitability. Overexpression of PDE4D1 in GT1 cells lowered cAMP levels, decreased the number of spontaneous Ca2+ oscillations, and decreased the frequency of GnRH pulses (7). Expression of PDE4D1 in GT1 cells also blocked the forskolin-induced increase in Ca2+ oscillations.

    The large decrease in LH pulse frequency observed in castrated GPR-4 rats was consistent with the findings in GT1 cells of lowered GnRH neuron excitability and decreased intrinsic release of GnRH pulses. The lowered excitability of the GnRH neurons in 4-month-old GPR-4 females also resulted in either complete or partial inhibition (64%) of the ovulatory surge of LH. The decreased amplitude of the ovulatory surge of LH was consistent with the hypothesis that the GnRH neurons were less responsive to stimulatory afferent inputs.

    Preliminary observations suggested that there was an age-related decrease in fertility of GPR-4 females. At 2.5 months of age, 80% of females produced litters, whereas at 6 months of age, no litters were observed. Interestingly, the ovaries of the nonovulating 4-month-old GPR-4 females contained large numbers of cystic follicles. These observations led us in the current study to detail the age-related loss of fertility in 3-, 6-, and 9-month-old GPR-4 females. We asked whether the appearance of the polycystic ovarian phenotype and loss of fertility correlated with changes in circulating levels of the reproductive hormones: LH, FSH, estradiol, androstenedione, and testosterone. Given that infertility has been linked to obesity, we tested whether infertility in the GPR-4 rat was correlated with obesity and changes in the levels of IGF-I, insulin, and leptin, hormones involved in the regulation of food intake and metabolism. Because we observed that 9-month-old GPR-4 rats developed abdominal obesity, we also asked whether the decreased excitability in GnRH neurons was associated with abnormalities in the regulation of caloric intake when 2-month-old females were provided with a high-energy diet (HED) for 24 d.

    Materials and Methods

    Animals: GPR-4 rats

    Expression of PDE4D1 was specifically targeted to GnRH neurons in the GPR-4 rat transgenic line using the promoter/enhancer regions of the rat GnRH gene. The transgene consisted of 3 kb of the promoter/enhancer region of the rat GnRH gene inserted upstream of the B intron of the rabbit -globin gene. The PDE4D1 cDNA was inserted downstream of the B intron followed by the human GH polyadenylation sequence. The GPR-4 transgenic line was established as previously described (2).

    For breeding, 2- to 3-month-old female or male GPR-4 transgenic rats were mated either with wild-type (WT) Sprague Dawley rats (the same background on which the GPR-4 rats were produced) or with GPR-4 rats. At weaning, rats were genotyped by tail blot analysis. The transgenic rats used were either heterozygous or homozygous for the transgene. There were no significant differences in the reproductive or metabolic variables measured in this study between heterozygous and homozygous transgenic rats, and results from homozygous and heterozygous rats were, therefore, pooled. Nontransgenic littermates were used as WT controls. Animals were housed three per cage in a temperature-controlled room with a 12-h light/dark cycle and free access to food and water. Animal care and handling conformed to National Institutes of Health guidelines for animal research, and the experimental protocols were approved by the University of California, San Francisco Committee on Animal Research.

    Sample/tissue collection

    For age-related studies, three groups of female WT and GPR-4 littermate rats were used at 3 months (12 WT and 15 GPR-4), 6 months (13 WT and 14 GPR-4), and 9 months of age (16 WT and 24 GPR-4). All animals were decapitated in the morning, trunk blood was collected into heparin-coated tubes, and plasma was obtained for hormone measurements. Abdominal white adipose tissue (AB-WAT; mesenteric, perinephric and ovarian) and sc white adipose tissue (SC-WAT) were dissected out and weighed. Vaginal smears were collected immediately after decapitation to evaluate the stage of the estrous cycle.

    For measuring estradiol levels on the afternoon of proestrus, in five WT and five GPR-4 rats (4 months old) a right atrial catheter was inserted on the morning of proestrus under tribromoethanol anesthesia (2). This surgical procedure did not block the preovulatory surge of LH. Proestrus estradiol levels were measured in samples obtained every hour between 1200 and 1800 h from awake freely behaving animals.

    For feeding studies, 2-month-old female WT and GPR-4 rats were divided into two groups with access to either chow or HED ad libitum. Two pelleted diets were used. Purina Rodent Chow (diet 5008; 23.5% protein, 6.5% fat, and 49.4% carbohydrates; metabolizable energy, 3.31 kcal/g; Purina, St. Louis, MO) was used as the control diet (chow). Purified diet D12266B (18.5% protein, 15.6 fat, and 56.7% carbohydrate; metabolizable energy, 4.41 kcal/g; Research Diets, Inc., Brunswick, NJ) was used as HED for metabolic, endocrine, and energy balance phenotyping. Water was always available. Rats were singly housed in hanging cages for at least 5 d before each experiment in a temperature-controlled room with a 12-h light/dark cycle. Body weights and food intake were measured daily for 24 d. At the end of the experiment, AB-WAT and SC-WAT depots were collected, cleaned, and weighed. This experiment was repeated, and after statistical analysis showed that the results were not different, the results of the independent experiments were pooled with the following group sizes: WT-chow, n = 11; WT-HED, n = 9; GPR-4-chow, n = 15; GPR-4-HED, n = 15.

    Ovarian histology

    Ovaries were dissected, weighed, fixed in 4% paraformaldehyde, embedded in paraffin, serially sectioned at 8 μm, and stained with hematoxylin/eosin. The number of antral follicles, corpora lutea, and cystic follicles were counted in every fifth section according to criteria and methodology previously published (8).

    RIAs

    LH levels were determined by RIA (The National Hormone and Peptide Program, National Institute of Diabetes and Digestive and Kidney Diseases). LH and all of the other hormone measurements were made in a single assay. The sensitivity of the assay was 0.05 ng/ml, and the intraassay coefficient of variation was less than 9%.

    FSH levels were measured using a rat RIA kit (ICN Diagnostics, Irvine, CA) according to the manufacturer’s instructions. The sensitivity of the assay was 0.2 ng/ml, and the intraassay coefficient of variation was less than 4%.

    Estradiol, androstenedione, and testosterone levels were measured using RIA kits (ICN Diagnostics). The sensitivity of the assays was 0.2 pg/ml, 0.04 ng/ml, and 0.08 ng/ml, respectively. The intraassay coefficients of variation in the three assays were less than 5%.

    IGF-I levels were measured using a rat RIA kit (Diagnostic Systems Laboratories, Inc., Webster, TX). The sensitivity of the assay was 30 ng/ml, and the intraassay coefficient of variation was less than 6%.

    Insulin and leptin levels were measured using rat RIA kits (Linco, St. Charles, MO). The sensitivity of the assays was 0.1 and 0.5 ng/ml, respectively. The intraassay coefficients of variation in both assays were less than 4%.

    Evaluation of fertility

    Fertility was evaluated in additional groups of 3-, 6-, and 9-month-old female WT and GPR-4 rats (eight WT and eight GPR-4 rats per group). Daily vaginal smears were obtained for 10 d before females were placed with an experienced fertile male for 15 d. Fertility was defined by the percentage of animals producing litters and litter size.

    Statistical analysis

    Data were expressed as mean ± SEM, and statistical differences were determined using ANOVA two-way analysis, corrected for repeated measures in feeding studies. Mann-Whitney U test or Fisher’s projected least significant difference were used to test post hoc significance.

    Results

    Age-related changes in fertility and ovarian histology

    At 3 months of age, all GPR-4 and WT rats had normal 4- to 5-d estrous cycles as determined by vaginal cytology (Table 1). However, at 6 months of age, 75% of GPR-4 rats persistently showed a characteristic vaginal smear consisting of both enlarged epithelial cells and partially cornified cells. All of the 6-month-old WT rats exhibited normal cycles. At 9 months of age, all GPR-4 rats persistently showed a true cornified vaginal smear, whereas 88% of WT rats exhibited the normal cyclical pattern of vaginal smears.

    Changes in fertility paralleled the changes seen in vaginal smears. At 3 months of age, 75% of GPR-4 rats produced litters compared with 100% of WT rats when housed with a fertile male for 15 d. Litter size did not vary between transgenic and WT rats. At 6 months of age, only 25% of GPR-4 females were fertile compared with 88% of WT rats. Furthermore, the 6-month-old fertile GPR-4 rats produced smaller litters. No litters were observed in 9-month-old GPR-4 rats, although 75% of WT rats were fertile. The decreased fertility closely correlated with changes in ovarian weight and histology.

    At all ages, ovarian weight was decreased in GPR-4 females relative to WT females (Table 2). However, at 6 months of age, the decrease was more pronounced than in 3-month-old animals, and no further decrease was observed at 9 months of age. The decrease in ovarian weight correlated with a decrease in the number of corpora lutea observed in histological sections (Fig. 1). The dramatic decrease in the number of corpora lutea in GPR-4 females at 6 and 9 months of age demonstrated the almost complete inhibition of ovulation. No significant changes in ovarian weight or the number of corpora lutea were seen in 3-, 6-, and 9-month-old WT females.

    Follicular development was assessed by counting the number of antral follicles. There were no differences in the number of antral follicles found in GPR-4 females at various ages or the number found between GPR-4 and WT females at any age. However, there were large increases in the number of cystic follicles seen at 6 and 9 months of age in the GPR-4 females. These observations are consistent with the idea that follicular development was normal through the formation of large antral follicles but was blocked around the time of ovulation.

    Age-related changes in reproductive hormones

    Gonadotropin levels are known to be key indexes of fertility and follicular development. Mean morning plasma LH levels were significantly lower (P < 0.01) in 3-, 6-, and 9-month-old GPR-4 rats compared with WT rats (Fig. 2). LH levels were decreased by 35–38% at different ages, consistent with the magnitude of decrease previously reported in LH levels in GPR-4 rats (2). Morning LH levels in GPR-4 and WT rats did not differ at 3, 6, or 9 months of age. No significant differences were found in plasma FSH levels between GPR-4 and WT rats at various ages. FSH levels tended to increase with age in both groups of animals, but the changes were not statistically significant.

    The mean plasma estradiol levels did not differ between GPR-4 and WT females of 3, 6, and 9 months of age, although estradiol levels were consistently higher in GPR-4 rats of all ages. Estradiol levels tended to increase with age in both groups of animals but the changes were not statistically significant. We also measured estradiol levels throughout the afternoon of proestrus in cycling 4-month-old WT (n = 5) and GPR-4 rats (n = 5); no significant difference was seen between the two groups on proestrus (WT = 36.3 ± 4 pg/ml and GPR-4 = 39.8 ± 6 pg/ml).

    Plasma androstenedione concentrations did not differ between GPR-4 and WT females at 3, 6, and 9 months of age. Also, no changes in androstenedione levels were observed at various ages in WT or GPR-4 females. Levels of testosterone in the plasma were undetectable in all of the samples studied (data not shown).

    Age-related changes in body weight and fat content

    We also carried out studies to determine whether GPR-4 and WT females showed any changes in body weight or fat content with age. As would be expected, body weight increased with age in the WT and GPR-4 females (Fig. 3). However, body weight was not different between GPR-4 and WT rats at any age.

    In the absence of a difference in body weight, 9-month-old GPR-4 females had heavier fat depots compared with 9-month-old WT controls. Nine-month-old GPR-4 females had significantly heavier AB-WAT and SC-WAT than 9-month-old WT females or 3- and 6-month-old GPR-4 females. Although the increases in AB-WAT and SC-WAT in 9-month-old GPR-4 rats compared with WT control seem small relative to body weight, they represented a 1.4- and 2.2-fold increase in the fat to body weight ratio, respectively. Also, both GPR-4 and WT rats had significant increases in AB-WAT, but not SC-WAT, at 6 months of age compared with 3-month-old rats

    Age-related changes in metabolic hormones

    The increase in fat content in 9-month-old GPR-4 females correlated closely with changes in the circulating levels of insulin and leptin. Insulin and leptin levels in 9-month-old GPR-4 females were significantly increased compared with WT controls or to 3- and 6-month-old GPR-4 rats (Fig. 4). However, no differences in insulin and leptin levels were seen between the two groups in 3- or 6-month-old females. Also, no significant differences were seen between IGF-I levels in the GPR-4 and WT females at any age. In both WT and GPR-4 females, IGF-I levels decreased at 6 months of age and remained low at 9 months of age.

    Effect of HED on obesity

    In both WT and GPR-4 females that were fed the HED, an acute increase in daily caloric intake, hyperphagia, was observed (Fig. 5A). However, whereas the WT rats restored ingestion of total calories to normal after 2 d, GPR-4 rats remained hyperphagic for the duration of the experiment. The increased food consumption in the GPR-4 rats on the HED was associated with a significant increase in the rate of body weight gain (Fig. 5B). Both the WT and GPR-4 rats showed significant increases in the amounts of sc fat; however, only the GPR-4 females on HED showed an increase in the metabolically important abdominal fat (Fig. 6A). The mean caloric efficiency (body weight gain grams per kilocalories ingested per day) on chow was not different between the WT (0.022 ± 0.002) and GPR-4 rats (0.022 ± 0.002). The mean caloric efficiency of GPR-4 (0.034 ± 0.002) but not WT rats (0.022 ± 0.003) on a HED was significantly increased (P < 0.05; Mann-Whitney U test). This finding indicates that although the level of food consumption was increased in the GPR-4, the rate of metabolism was decreased.

    Effect of HED on reproductive and metabolic hormones

    As previously observed, GPR-4 females had lower plasma LH levels and ovarian weights than WT females (Table 3). Diet had no effect on the decreased LH levels and ovarian weight in the GPR-4 females. Fifty percent of the GPR-4 rats on either diet were nonovulatory, based on absence of corpora lutea. Plasma FSH and estradiol levels were similar in WT and GPR-4 rats on chow or HED. The only significant effect of HED on reproductive hormones was an increase in circulating levels of androstenedione in GPR-4 rats.

    GPR-4 females had significantly (P < 0.05; Mann-Whitney U test) higher plasma corticosterone levels than WT rats on chow (21 ± 7 vs. 14 ± 3 μg/dl, respectively) and HED (21 ± 5 vs. 10 ± 6 μg/dl, respectively). HED did not affect circulating insulin and leptin concentrations in WT rats. However, GPR-4 females on HED had significantly higher leptin and insulin levels on HED (Fig. 6).

    Discussion

    These studies clearly define the premature loss of fertility in the GPR-4 transgenic female rats compared with WT littermate females. The almost complete loss of fertility at 6 months of age correlated closely with the cessation of ovulation and the development of ovarian cysts. However, the onset of obesity, defined as an excess of both AB-WAT and SC-WAT relative to body weight, and changes in insulin and leptin were not seen until 9 months of age. Interestingly, 2-month-old GPR-4 females, when placed on HED, overeat and develop abdominal obesity. A number of questions arise on how expression of the PDE4D1 transgene in GnRH neurons resulted in the loss of fertility, the development of obesity at 9 months of age, and the dysregulation of energy balance when fed HED at 2 months of age.

    It appears that the central event in the loss of cyclicity and fertility was inhibition of ovulation. The lack of ovulation in 6-month-old GPR-4 females was confirmed by the decreased number of corpora lutea. We hypothesize that the inhibition of ovulation was the result of inhibition of the normal GnRH-induced ovulatory LH surge resulting from expression of PDE4D1 in GnRH neurons. In this study, decreased LH levels were observed in morning LH levels. In an earlier study, only half of 4-month-old GPR-4 females had normal estrous cycles. In these 4-month-old cycling GPR-4 rats, the ovulatory surge of LH was decreased by 64% (2). The decrease in the amplitude of the ovulatory surge was not caused by a decrease in elevated estradiol levels on the afternoon of proestrus. Estradiol levels in cycling 4-month-old GPR-4 and WT females were the same on the afternoon of proestrus. This observation leads us to hypothesize that there is a decrease in the positive feedback of estradiol on GnRH neurons in the GPR-4 female. In middle-aged rats (8–14 months old), before the loss of cyclicity, the positive feedback action of estradiol is diminished, whereas in aged acyclic female rats it is lost (9).

    Expression of PDE4D1 in GT1 GnRH neurons resulted in a decrease in the frequency of intrinsic GnRH pulses that were associated with a decrease in excitability of the neurons (7). Also there was a decrease in the ability of the PDE4D1-expressing GT1 cells to respond to the neurotransmitter dopamine or pharmacological agents that increase cAMP levels. We hypothesize that, in a similar fashion, the decreased pulsatile release of LH in castrated GPR-4 rats and the decreased LH ovulatory surge in cycling GPR-4 rats were the result of decreased excitability of the GnRH neurons. It follows from this argument that whether the positive feedback actions of estradiol are mediated via receptors in GnRH neurons or receptors in neurons that alter the activity of GnRH neurons through synaptic interactions, there would be a decrease in the positive feedback actions of estradiol in GPR-4 rats. Future studies will determine whether the amplitude of the preovulatory surge decreases progressively with age in the GPR-4 rats, and whether the decrease is due to a decrease in the positive feedback response to elevated estradiol.

    A second potential explanation for the loss of ovulation would be the inhibition of follicular development by lowered LH levels or the inability of a decreased preovulatory LH surge to induce ovulation of Graafian follicles. The number of antral follicles in 3-, 6-, and 9-month-old GPR-4 females were the same as in WT females. Clearly, there was no decrease in the number of antral follicles in GPR-4 females at 9 months of age. Therefore, it appears that FSH and LH levels were sufficient to support follicular development. Consistent with this idea, morning estradiol levels were not decreased in 3-, 6-, or 9-month-old GPR-4 females. Interestingly, in 4-month-old cycling GPR-4 females, the number of ova ovulated were unchanged compared with WT controls (2). This finding is consistent with the conclusion that the number of large antral follicles available for ovulation is unchanged in the GPR-4 females. We conclude that decreased LH levels and unchanged FSH levels were sufficient for follicular development and steroidogenesis in the GPR-4 female.

    As mentioned, the diminished preovulatory LH surge in 4-month-old GPR-4 rats was sufficient to induce ovulation of the same number of follicles seen in WT females. However, the lowered LH surge may not have been sufficient to induce ovulation of Graafian follicles in the 6-month-old GPR-4 rats. A decrease in the sensitivity of ovarian follicles to LH stimulation of ovulation was observed in aging rats (10). By 6 months of age, the possibility exists that the remaining follicles in the ovaries of GPR-4 females are less responsive to the ovulatory action of LH. This question could be tested in 6-month-old GPR-4 females by attempting to rescue ovulation through administration of LH.

    Large numbers of cystic follicles were observed in the ovaries of 6-month-old GPR-4 females. Cystic follicles were defined as large follicles that did not contain an ovum and were lined by only one or two layers of granulosa cells. How the blockade of ovulation relates to the development of cystic follicles is not clear. The increased number of cystic follicles closely paralleled the decrease in the number of corpora lutea. This observation is reminiscent of findings in several mammals in which blockade of ovulation by a variety of mechanisms results in development of cystic follicles. In cattle, inhibition of the ovulatory surge of LH by immuno-neutralization of estradiol resulted in the formation of ovarian cysts (11). The ovaries of anovulatory middle-aged rats contain large numbers of follicular cysts (12). An injection of a single large dose of estradiol valerate to young adult rats resulted in anovulation and the appearance of polycystic ovaries (13). Transgenic mice overexpressing LH ovulate infrequently and have polycystic ovaries (14).

    There was a clear age-related metabolic phenotype in GPR-4 females. At 9 months of age there was a significant increase in both abdominal and sc fat in GPR-4 rats. The increased obesity was closely correlated with an increase in plasma insulin and leptin levels. Because the metabolic changes occurred several months after the cessation of ovulation and development of polycystic ovaries seen at 6 months of age, it is difficult to argue that these play a significant role in the loss of fertility. Also, metabolic changes could not be ascribed to changes in steroid hormone levels because no obvious age-related changes in reproductive steroids were seen in GPR-4 rats relative to WT controls. Basal plasma estradiol and androstenedione levels in WT and GPR-4 rats did not differ at 3, 6, or 9 months of age. It is possible that small differences in estradiol levels in a specific day of the estrous cycle could have been missed because the rats were killed in the morning of random days of the estrous cycle. However, from individual hormonal data and vaginal smears collected on the day the rats were killed, there did not appear to be large differences in estradiol levels measured across the cycle. Also, no differences were observed in the preovulatory elevations in estradiol in 4-month-old WT and GPR-4 females.

    The obesity phenotype in the 9-month-old GPR-4 females led us to ask how young GPR-4 females would respond to the challenge of a HED. The experiments were performed in 2-month-old females to eliminate the effects of age-related changes in hormone levels or aging of the neuroendocrine mechanisms regulating reproduction and food intake. GPR-4 females on the HED developed hyperphagia, abdominal obesity, and increased caloric efficiency. Initially both the WT and GPR-4 rats on HED overate; however, WT females rapidly acclimated to the HED and restored caloric consumption to the same levels as WT females on chow. The normal response to increased energy intake is to increase energy expenditure, thus decreasing caloric efficiency and maintaining constant energy stores. Therefore, the increased caloric efficiency observed in GPR-4 females was unexpected and helps to define the phenotypic change in energy management. Caloric efficiency (calculated as body weight gain/energy ingested/day) comprises, in addition to altered metabolic rate, altered nutrient absorption and voluntary activity; thus, it is quite possible that the increased caloric efficiency observed in GPR-4 rats included increased nutrient absorption and decreased voluntary activity. Clearly the mechanisms regulating food intake were altered in the GPR-4 females. Another parallel to the finding in the 9-month-old GPR-4 obese females was that young GPR-4 females on the HED had elevated levels of both insulin and leptin. Two hypotheses that may explain these findings are that: decreased GnRH release leads to alterations in the hormone milieu that participates in regulating food intake, and, decreased excitability of GnRH neurons in the GPR-4 females alters the bias of the neuronal circuitries that regulate energy balance.

    Estradiol affects both food intake and energy expenditure. Female rats that are ovariectomized increase food intake and body weight, the latter mainly due to increases in fat weight, and decreases in voluntary physical activity and thermogenesis. These effects are reversed by estradiol administration. Moreover, the fluctuations in sex hormones in female rats (and women) during the estrous (or menstrual) cycles correlate with fluctuations in energy balance (15). However, no alterations in estradiol levels were observed between WT and GPR-4 females on HED or chow at the time they were killed. The possibility that estradiol levels could have varied over the course of the experiment seems unlikely from the data presented above. No differences were seen in estradiol levels in morning samples or during the preovulatory increase on proestrus. Although we observed an increase in androstenedione levels in the GPR-4 females on the HED, there is no reason to believe this change affected food intake, because administration of androstenedione to ovariectomized female rats has no effect on food intake or body weight (16).

    Both the adipocyte-derived hormone leptin and pancreatic insulin act as adiposity signals to the hypothalamus as part of negative feedback loops that sense the caloric stores of an animal and accordingly regulate energy balance (17). Both hormones act centrally to inhibit feeding and increase energy expenditure (18). Our findings of obesity despite elevated insulin and leptin levels in the 9-month-old GPR-4 female rats eating chow and in the 2-month-old GPR-4 females on HED, suggests that the GPR-4 females are insulin and leptin resistant. Resistance to insulin and leptin are common features in obesity (19). This observation leads us to our second hypothesis that GnRH neurons may participate in the neuronal circuitry regulating food intake and that lowered excitability of GnRH neurons is associated with increased food intake. Perhaps the resistance to the anorexic actions of insulin and leptin or other metabolic signals is related to the transgene-induced decrease in the excitability of GnRH neurons.

    Whether the GPR-4 rat is relevant to the understanding of the development of polycystic ovarian syndrome (PCOS) in women is still to be fully understood. Major similarities include inhibition of reproductive cyclicity, the development of large polycystic ovaries, and an association with obesity and insulin and leptin resistance (20). Elevated androgen levels are a hallmark of PCOS (21), and although the androstenedione levels in GPR-4 females eating chow were unaltered, GPR-4 females on a HED had elevated levels of androstenedione. In addition, P450c17 hydroxylase immunostaining was increased in GPR-4 females (22), as is P450c17 hydroxylase activity in woman with PCOS (23). Further experiments are necessary to assess the androgen-producing capacity of the ovary in GPR-4 females.

    However, major differences exist in the gonadotropin levels and pulsatile GnRH release in GPR-4 rats and in women with PCOS. LH levels are elevated and the LH pulse frequency increases in women with PCOS (24), whereas both are decreased in the GPR-4 rat. FSH levels are lower in women with PCOS (25) and unchanged in GPR-4 rats. The data in the GPR-4 female are consistent with decreased excitability of GnRH neurons; however, the observation in women with PCOS of an increased LH pulse frequency suggests that different mechanisms lead to polycystic ovaries in women and GPR-4 females.

    In summary, we showed that transgenic GPR-4 female rats were characterized by a premature loss of fertility, anovulation, development of polycystic ovaries, and dysregulation of the mechanisms regulating energy balance. The loss of fertility closely correlated with decreased LH levels and the interruption of ovulation occurred before the onset of obesity. The cause of the onset of obesity in 9-month-old GPR-4 female rats appears to be related to an eating disorder. The onset of obesity was not correlated with changes in basal plasma estradiol but rather with elevated levels of leptin and insulin, consistent with resistance to the anorexic action of these hormones. Overeating and the resistance to insulin and leptin may reflect a role for GnRH neurons in the neural circuitry regulating food intake. Further studies are necessary to understand the possible role of GnRH neurons in the development of obesity.

    Acknowledgments

    We thank Segundi San Juan, Sotara Manalo, and Dr. Norman Pecoraro for their skilled work on the project.

    Footnotes

    This work was supported by National Institutes of Health Grant HD 41996 (to R.I.W.).

    1 F.G. and S.E.l.F. contributed equally to this work.

    Abbreviations: AB-WAT, Abdominal white adipose tissue; HED, high-energy diet; PCOS, polycystic ovarian syndrome; SC-WAT, sc white adipose tissue; WT, wild type.

    References

    Knobil E 1980 The neuroendocrine control of the menstrual cycle. Recent Prog Horm Res 36:53–88

    Paruthiyil S, El Majdoubi M, Conti M, Weiner RI 2002 Phosphodiesterase expression targeted to gonadotropin-releasing hormone neurons inhibits luteinizing hormone pulses in transgenic rats. Proc Natl Acad Sci USA 99:17191–17196

    Martinez de la Escalera G, Gallo F, Choi AL, Weiner RI 1992 Dopaminergic regulation of the GT1 gonadotropin-releasing hormone (GnRH) neuronal cell lines: stimulation of GnRH release via D1-receptors positively coupled to adenylate cyclase. Endocrinology 131:2965–2971

    Martinez de la Escalera G, Choi AL, Weiner RI 1992 1-Adrenergic regulation of the GT1 gonadotropin-releasing hormone (GnRH) neuronal cell lines: stimulation of GnRH release via receptors positively coupled to adenylate cyclase. Endocrinology 131:1397–1402

    Martinez de la Escalera G, Choi AL, Weiner RI 1995 Signaling pathways involved in GnRH secretion in GT1 cells. Neuroendocrinology 61:310–317

    Charles A, Weiner R, Costantin J 2001 cAMP modulates the excitability of immortalized hypothalamic (GT1) neurons via a cyclic nucleotide gated channel. Mol Endocrinol 15:997–1009

    Yoshida H, Beltran-Parrazal L, Butler P, Conti M, Charles AC, Weiner RI 2003 Lowering cyclic adenosine-3',5'-monophosphate (cAMP) levels by expression of a cAMP-specific phosphodiesterase decreases intrinsic pulsatile gonadotropin-releasing hormone secretion from GT1 cells. Mol Endocrinol 17:1982–1990

    Lara HE, Dissen GA, Leyton V, Paredes A, Fuenzalida H, Fiedler JL, Ojeda SR 2000 An increased intraovarian synthesis of nerve growth factor and its low affinity receptor is a principal component of steroid-induced polycystic ovary in the rat. Endocrinology 141:1059–1072

    Wise PM 1984 Estradiol-induced daily luteinizing hormone and prolactin surges in young and middle-aged rats: correlations with age-related changes in pituitary responsiveness and catecholamine turnover rates in microdissected brain areas. Endocrinology 115:801–809

    Tain CF, Goh VH, Ng SC 2000 Effects of hyperstimulation with gonadotrophins and age of females on oocytes and their metaphase II status in rats. Mol Reprod Dev 55:104–108

    Kaneko H, Todoroki J, Noguchi J, Kikuchi K, Mizoshita K, Kubota C, Yamakuchi H 2002 Perturbation of estradiol-feedback control of luteinizing hormone secretion by immunoneutralization induces development of follicular cysts in cattle. Biol Reprod 67:1840–1845

    Lu KH, Hopper BR, Vargo TM, Yen SS 1979 Chronological changes in sex steroid, gonadotropin and prolactin secretions in aging female rats displaying different reproductive states. Biol Reprod 21:193–203

    Schulster A, Farookhi R, Brawer JR 1984 Polycystic ovarian condition in estradiol valerate-treated rats: spontaneous changes in characteristic endocrine features. Biol Reprod 31:587–593

    Risma KA, Clay CM, Nett TM, Wagner T, Yun J, Nilson JH 1995 Targeted overexpression of luteinizing hormone in transgenic mice leads to infertility, polycystic ovaries, and ovarian tumors. Proc Natl Acad Sci USA 92:1322–1326

    Wade GN, Gray JM 1979 Gonadal effects on food intake and adiposity: a metabolic hypothesis. Physiol Behav 22:583–593

    Rowland DL, Perrings TS, Thommes JA 1980 Comparison of androgenic effects on food intake and body weight in adult rats. Physiol Behav 24:205–209

    Figlewicz DP 2003 Adiposity signals and food reward: expanding the CNS roles of insulin and leptin. Am J Physiol Regul Integr Comp Physiol 284:R882–R892

    Air EL, Benoit SC, Clegg DJ, Seeley RJ, Woods SC 2002 Insulin and leptin combine additively to reduce food intake and body weight in rats. Endocrinology 143:2449–2452

    Cases JA, Barzilai N 2000 The regulation of body fat distribution and the modulation of insulin action. Int J Obes Relat Metab Disord 24(Suppl 4):S63–S66

    Gambineri A, Pelusi C, Vicennati V, Pagotto U, Pasquali R 2002 Obesity and the polycystic ovary syndrome. Int J Obes Relat Metab Disord 26:883–896

    Yen SS 1980 The polycystic ovary syndrome. Clin Endocrinol (Oxf) 12:177–207

    El Majdoubi M, Gomez F, La Fleur S, Manola S, Mellon SH, Dallman MF, Weiner RI, Age-related development of anovulation and cystic ovarian follicles in the GPR-4 transgenic rat. Program of the 85th Annual Meeting of The Endocrine Society, Philadelphia, PA, 2003, p 485 (Abstract P3-45)

    Ehrmann DA, Barnes RB, Rosenfield RL 1995 Polycystic ovary syndrome as a form of functional ovarian hyperandrogenism due to dysregulation of androgen secretion. Endocr Rev 16:322–353

    Morales AJ, Laughlin GA, Butzow T, Maheshwari H, Baumann G, Yen SS 1996 Insulin, somatotropic, and luteinizing hormone axes in lean and obese women with polycystic ovary syndrome: common and distinct features. J Clin Endocrinol Metab 81:2854–2864

    Yen SS, Vela P, Rankin J 1970 Inappropriate secretion of follicle-stimulating hormone and luteinizing hormone in polycystic ovarian disease. J Clin Endocrinol Metab 30:435–442(Francisca Gomez1, Susanne)