|
|
||||||||
Regular Article |
a Departments of Medicine, Cell Biology, and Psychiatric Medicine, University of Virginia, Charlottesville, Virginia 22908
| ABSTRACT |
|---|
|
|
|---|
gonadotropin-releasing hormone, leptin, ovulatory cycle, serotonin
| INTRODUCTION |
|---|
|
|
|---|
The serotonergic system plays a role in both food intake [17] and GnRH release (reviewed in [18]) and is one candidate mediator of leptin action. Specifically, serotonin (5HT) 1 and 2 receptor subtypes are involved in the regulation of food intake; antagonizing these receptors with metergoline, a type 1/2/7 5HT-receptor antagonist, blocks the anorectic effects of fluoxetine [19]. Anatomical evidence exists for a link between the serotonergic and GnRH systems: serotonergic neurons terminate directly on GnRH neurons in the medial preoptic area in rats [20, 21] and are leptin-responsive [22]. Furthermore, leptin-receptor message is expressed in brain stem serotonergic neurons, suggesting that they are direct targets of leptin action and may mediate some of the effects of leptin on feeding and reproduction [23]. Several reports of functional interactions between leptin and serotonergic pathways have appeared, although the nature of this relationship is controversial. Leptin and 5HT exerted similar effects on satiety, antagonizing neuropeptide Y action [24]. The 5HT precursor 5-hydroxytryptophan induced a dose-dependent increase in serum leptin levels in mice [25]; in contrast, fluoxetine, which inhibits 5HT reuptake, thereby increasing serotonergic tone, decreased leptin in lean and obese Zucker rats [26]. In wild-type mice, leptin caused a dose-dependent increase in levels and turnover of 5HT, suggesting activation of these neurons by leptin [27]. Similarly, in ob/ob mice, leptin doses sufficient to decrease food intake and body weight also increased brain stem and hypothalamic 5HT concentrations [28].
To investigate the role of the serotonergic system as a mediator of leptin action on the reproductive axis, 3 hypotheses were tested. First, increasing serotonergic tone by selectively blocking 5HT reuptake will rescue estrous cycles of fasted mice. Second, this action, as well as the ability of leptin to rescue estrous cycles, will be blocked by antagonizing serotonergic receptor subtypes involved in feeding behavior [19, 29]. Third, blocking 5HT reuptake will induce weight loss and rescue reproduction in leptin-deficient ob/ob and leptin receptor-deficient db/db mice.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Female C57Bl6-J mice (810 wk of age) were obtained from Jackson Laboratories (Bar Harbor, ME) or Hilltop (Scottdale, PA) and were housed individually during experiments. Mice were maintained on standard rodent chow (7201; Harlan, Madison, WI) and water ad libitum except for one 48-h fast, during which only water ad libitum was allowed. Female ob/ob and db/db mice (810 wk of age; Jackson Laboratories) were housed in breeding cages with 3 females and 1 sexually mature, normal male from an in-house breeding colony per cage. These mice were maintained on breeder chow (8656; Harlan) and water ad libitum. All animals were held on a 14L:10D photoperiod, with lights-on at 0500 h Eastern Standard Time. All procedures were approved by the Animal Care and Use Committee of the University of Virginia and conducted in accordance with the National Research Council publication Guide for Care and Use of Laboratory Animals.
Procedure
Experiment 1: effect of manipulating 5HT action on estrous cyclicity in fasted mice Body weight was recorded daily. To monitor estrous cycles, daily vaginal smears were taken for 3 consecutive cycles. Mice with regular 4- to 5-day cycles were housed individually and randomly placed in 1 of 8 treatment groups. Beginning on Diestrous Day 1, animals were fasted for 48 h and treated with one of the following: saline twice daily i.p. (bid i.p.; n = 6, 0.1 ml), recombinant mouse leptin (n = 5, 0.1 mg/kg bid i.p.; Sigma, St. Louis, MO), fluoxetine (n = 6, 32 mg/kg once s.c. at the start of fast; Eli Lilly, Indianapolis, IN), fluoxetine as above plus the 5HT 1/2/7 receptor-antagonist metergoline (n = 5, 1 mg/kg bid i.p.; Sigma), leptin as above plus metergoline at 1 of 2 doses (1 or 2 mg/kg bid i.p., n = 7 and 6, respectively), or metergoline alone at these same doses (n = 7 and 4, respectively). Vaginal smears and body weight were monitored during fasting and until the return to estrus following treatment. Days to return to estrus (i.e., length of the treatment estrous cycle), food intake, weight loss during fast, weight gain on refeeding, and body weight at return to estrus were compared among groups by ANOVA followed by post-hoc pairwise analysis with Tukey-Kramer, Student-Newman-Keuls, and Fisher protected least significant difference tests when appropriate. All data are reported as the mean ± SEM.
Experiment 2: effect of blocking 5HT reuptake on reproduction in ob/ob and db/db mice A pilot study was performed on female ob/ob and db/db mice to assess tolerance for prolonged daily fluoxetine treatment. Because treatment was to be long term, a dose of fluoxetine lower than that used in experiment 1 was chosen initially (20 mg/kg i.p. for daily injections vs. 32 mg/kg i.p. for a single injection). Mice treated daily with 20 mg/kg i.p. lost weight; however, this regimen caused constipation in some mice. To avoid this problem during the main experiment, the frequency of injection was reduced to every other day (q2d), both because fluoxetine has a long half-life [30] and because the dose of fluoxetine was increased gradually to reach 20 mg/kg (see below).
Female ob/ob and db/db mice were treated q2d i.p. with fluoxetine (n = 6 each of ob/ob and db/db) or water vehicle (n = 6 each of ob/ob and db/db). Treatment was initiated at 5 mg/kg and increased every 8 days by 5 mg/kg until doses reached 20 mg/kg. Treatments continued for a total of 65 days (42 days on the final 20 mg/kg q2d dose). Body weight was recorded and mice checked for the presence of copulatory plugs at each treatment. Vaginal smears were taken to assess estrous cyclicity.
| RESULTS |
|---|
|
|
|---|
Estrous cycle lengths Estrous cyclicity was used as a marker to determine if various treatments affected the degree of reproductive inhibition caused by fasting; estrous cycle lengths are shown in Figure 1. Estrous cycles of saline-treated nonfasted mice were 4.5 ± 0.2 days. Cycles of saline-treated fasted mice were significantly longer (10.2 ± 0.5 days, P < 0.0001), with the increase occurring in days spent in diestrus. As previously shown [3], treatment of fasted mice with leptin prevented this cycle lengthening (4.7 ± 0.7 days, P = 0.410 vs. fed controls). Of interest, fluoxetine treatment rescued estrous cycles with an efficacy similar to that of leptin, supporting the first hypothesis that these agents have similar actions on reproduction in a state of negative energy balance (cycle length for fluoxetine, 4.7 ± 0.6 days; P = 0.230 vs. fed controls, P < 0.005 vs. saline-treated fasted animals).
|
To further investigate whether the similar effects of leptin and fluoxetine on estrous cyclicity were achieved through parallel pathways or the same pathway, mice were treated with the 5HT 1/2/7 receptor-antagonist metergoline during fasting. Coadministration of metergoline with fluoxetine completely blocked estrous cycle rescue by fluoxetine (9.4 ± 0.2 days, P < 0.0001 vs. fed controls). Interestingly, metergoline also blocked cycle rescue by leptin in a dose-dependent manner, suggesting that leptin and 5HT lie along the same pathway. Metergoline at 2 mg/kg effectively blocked leptin rescue (8.7 ± 0.9 days, P < 0.0005 vs. fed controls), whereas a lower metergoline dose (1 mg/kg) did not have a significant effect (5.4 ± 0.9 days, P = 0.190 vs. fed controls). Metergoline alone at either dose (1 or 2 mg/kg bid i.p.) did not potentiate or block the effects of fasting on cycle length (8.9 ± 0.8 and 8.3 ± 0.5 days, respectively; P < 0.0003 and P < 0.0001, respectively, vs. fed controls).
Food intake and body weight Food intake and body weight were monitored to determine if among-treatment differences in these parameters might have influenced reproductive responses. All fasted animals lost the same percentage of body weight during the 48-h fast (Fig. 2A) and had the same total body weight gain at 24 h following refeeding (Fig. 2B). Furthermore, all animals had the same total food intake during the 24 h after refeeding, regardless of treatment (Fig. 2C). Of interest, mice treated with fluoxetine, leptin, or leptin plus low-dose metergoline (1 mg/kg bid i.p.) returned to estrus at a body weight less than their prefast body weight, whereas those treated with saline, metergoline, fluoxetine plus metergoline, or leptin plus high-dose metergoline (2 mg/kg bid i.p.) did not return to estrus until reestablishing or surpassing their prefast body weight (Fig. 3). This supports the notion that both leptin and fluoxetine treatments provide or substitute for a metabolic signal that is missing in the other fasted animals until prefast body weight is reestablished.
|
|
Experiment 2
The above findings suggest that leptin rescues reproductive function in states of negative energy balance by activating serotonergic pathways. We thus tested a third hypothesis: increasing serotonergic tone in genetically leptin-deficient ob/ob mice or leptin receptor-deficient db/db mice will rescue reproductive activity in these otherwise infertile animals. Long-term administration of fluoxetine at the doses used in this study had no effect on body weight, estrous cyclicity, or fertility of ob/ob and db/db mice compared to controls, despite this treatment lasting longer than the leptin treatment needed to restore fertility to ob/ob mice [5]. The db/db animals remained in constant diestrus and the ob/ob animals in constant estrus, regardless of treatment (fluoxetine or vehicle control), throughout the experimental period. Accordingly, fertility was not rescued by this treatment, as assessed by the lack of estrous cycles, absence of copulatory plugs, and failure to establish pregnancy.
| DISCUSSION |
|---|
|
|
|---|
Leptin and fluoxetine exerted similar effects on reproductive function in the fasting model utilized for this study. To begin to distinguish whether this observation was due to leptin and 5HT pathways acting in series or in parallel, pharmacological blockade of the serotonergic pathway was used. The type 1/2/7 5HT receptor-antagonist metergoline was chosen because 5HT 1 and 2 receptor subtypes are involved in the regulation of food intake and because antagonizing these receptors blocks the anorectic effects of fluoxetine [19]. Metergoline successfully blocked the ability of both fluoxetine and leptin to rescue estrous cycles. This suggests that leptin and serotonergic neurons act within a common pathway to GnRH neurons, and that 5HT-receptor subtype 1, 2, and/or 7 is used for this communication. This finding supports and extends those of previous reports. Specifically, although several reports of positive reproductive effects of leptin have appeared [512, 15, 16], to our knowledge, no evidence exists that GnRH neurons express any form of the leptin receptor [15, 16]. Serotonergic neurons express leptin receptors [23], are functionally down-regulated in leptin-deficient animals [22], and project directly to GnRH cell bodies [20, 21]. Taken together, these data suggest an anatomical pathway for leptin action on GnRH neurons, specifically that leptin signals are received by serotonergic neurons, which, in turn, communicate with GnRH neurons to regulate GnRH release. The existence of such a communication pathway does not rule out the possibility that reproductive rescue by leptin may also be mediated by direct stimulation of GnRH neurons and/or pituitary gonadotropes. Indeed, leptin has been shown to stimulate directly GnRH release from immortalized GnRH cell lines [14] as well as LH and FSH release from anterior pituitary explants [12]
To test the proposed pathway, leptin-deficient ob/ob and leptin receptor-deficient db/db mice were treated with fluoxetine and their reproductive activity assessed. No apparent rescue of estrous cycles, mating, or pregnancy was observed in these mice. This failure could be a result of several factors. It is possible that leptin and 5HT do not, in fact, work in series to regulate reproduction, although the blockade of leptin action by metergoline in experiment 1 argues against this hypothesis. Another possible explanation is that serotonergic systems are down-regulated in ob/ob (and possibly in db/db) mice. Consistent with this, it was recently reported that 5HT reuptake transporter mRNA levels are decreased in dorsal raphe nuclei of ob/ob mice compared to lean control animals, and that the behavior of these mice is consistent with decreased serotonergic activity [22]. The raphe nuclei comprise a major serotonergic relay station and are central neural regulators of feeding behavior [31] projecting to the hypothalamus [32]. Such a down-regulation of serotonergic systems in ob/ob (and possibly in db/db) mice due to leptin deprivation would not be expected to be overcome by inhibition of 5HT reuptake alone, and it may account for the failure of fluoxetine to decrease body weight or to rescue fertility in this study. A possible mechanistic explanation for the rescue of fertility of ob/ob mice by leptin [2] may, thus, be to increase the activity of dorsal raphe serotonergic neurons expressing leptin receptors and projecting to the hypothalamus.
Taken together, these data suggest a possible permissive role for serotonergic systems in the metabolic control of reproductive function. Specifically, serotonergic neurons may convey metabolic signals, such as leptin, to GnRH neurons in wild-type mice. Of interest, hypothalamic 5HT levels are increased by feeding [33] and decreased by food restriction [34, 35], findings that are consistent with the hypothesis that 5HT mediates communication between energy balance and reproductive function. Defining the pathways mediating 5HT action on the neuroendocrine axis may help to explain not only the metabolic control of reproduction but also the cause of numerous human disorders affecting both food intake and reproductive function, including Prader-Willi syndrome and anorexia nervosa. Of interest in this regard, women with Prader-Willi syndrome, a disease marked by hypothalamic hypogonadism, morbid obesity, and compulsive food-seeking behavior, established first menses 6 mo after initiation of treatment with fluoxetine [36]. Many women with anorexia nervosa are secondarily amenorrheic, exhibiting pulsatile LH secretion resembling that of the prepubertal stage; interestingly, these women also have abnormally low serotonergic tone [37, 38]. These clinical reports, together with the present data, demonstrate the importance of serotonergic and other pathways communicating metabolic status to the reproductive neuroendocrine axis, and they suggest new investigation and treatment strategies.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
1 Supported by NSF IBN9805923 (to S.M.M.). S.D.S. supported through the Neuroscience Training Program by funds from the School of Medicine. ![]()
2 Correspondence: Suzanne M. Moenter, Departments of Medicine and Cell Biology, University of Virginia, P.O. Box 800578, Charlottesville, VA 22908. FAX: 434 982 0088; smm4n{at}virginia.edu ![]()
Accepted: December 27, 2001.
Received: October 2, 2001.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
W. Huang, M. Acosta-Martinez, T. H. Horton, and J. E. Levine Fasting-induced suppression of LH secretion does not require activation of ATP-sensitive potassium channels Am J Physiol Endocrinol Metab, December 1, 2008; 295(6): E1439 - E1446. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. D. Sullivan and S. M. Moenter {gamma}-Aminobutyric Acid Neurons Integrate and Rapidly Transmit Permissive and Inhibitory Metabolic Cues to Gonadotropin-Releasing Hormone Neurons Endocrinology, March 1, 2004; 145(3): 1194 - 1202. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. D. Sullivan, R. A. DeFazio, and S. M. Moenter Metabolic Regulation of Fertility through Presynaptic and Postsynaptic Signaling to Gonadotropin-Releasing Hormone Neurons J. Neurosci., September 17, 2003; 23(24): 8578 - 8585. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |