| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Neuroscience Graduate Program and Department of Biology, University of Virginia, Charlottesville, Virginia 22903
ABSTRACT
Undernutrition has well-established effects on female reproduction. Here we describe the effects of food restriction on aspects of the hypothalamic-pituitary-gonadal (HPG) axis in the female musk shrew. We determined that acute re-feeding reverses deficits brought on by food restriction. Two days of food restriction led to an increase in proGnRH immunoreactive cells in the preoptic area relative to ad libitum-fed controls (AL). This increase was reversed by 90 min of ad libitum feeding in the re-fed females (RF). In addition, food-restricted (FR) females had significantly greater GnRH content in the median eminence than either the AL or RF females. After GnRH was administered, the majority of females in all food conditions ovulated, yet the FR females had significantly fewer corpora lutea than either the AL or RF animals. These data show that food restriction impairs HPG axis function in female musk shrews, and that some of these impairments can be rapidly reversed by acute re-feeding.
corpus luteum, follicular development, GnRH, ovulation, ovulatory cycle
INTRODUCTION
Food deprivation leads to infertility in many female mammals including humans, hamsters, mice, rats, and sheep [1–5]. Moreover, food deprivation induces physiological changes at several levels of the hypothalamic-pituitary-gonadal (HPG) axis. Food restriction affects GnRH neuron immunoreactivity. Lambs, growth restricted for 2 mo, have a larger proportion of GnRH immunoreactive (GnRH-ir) neurons in the medial-basal hypothalamus as compared to rapidly grown lambs of the same age [5]. Food deprivation also affects pituitary function. Female rats maintained at weaning weight for 31–38 days have a decreased LH pulse frequency and amplitude when compared to ad libitum fed rats [6]. Food restriction can increase pituitary sensitivity to GnRH. Food restricted hens have higher plasma LH than ad libitum fed controls after administration of chicken LHRH-I (cLHRH-I) [7]. Fasted male and female human subjects also have higher plasma concentrations of LH and FSH than nonfasted controls after GnRH injection [8]. Ovarian physiology and steroidogenesis are also affected by undernutrition. Schneider and Wade [2] reported that food deprivation on Days 1 and 2 of the estrous cycle retards follicular development and decreases plasma estradiol levels in hamsters [2].
In this study, we used the female musk shrew to examine the impact of undernutrition and brief reinstatement of food on the HPG axis. Musk shrews have several unique properties that make them a useful model species. First, they do not have an estrous cycle, and circulating steroid hormones are low prior to mating. Thus, nutritional manipulation can be implemented at any time without concern for the stage of the estrous cycle or variation in circulating steroid levels [9]. Second, musk shrews are induced ovulators and do not display a behavioral estrus that coincides with peaks in steroid hormones. Typically, they mate and subsequently ovulate whenever they interact with a male, despite the fact the testosterone and estradiol levels are low at this time [10]. This allows us to study direct effects of food restriction on behavior. Also, we can precisely control the onset of puberty (as defined by the first ovulation). Third, previous studies in our laboratory show that a food restriction schedule that is less severe, but similar in duration, to that used in rodents is sufficient to disrupt mating behavior in female musk shrews [11, 12]. In addition, this behavioral deficit is reversed after only 90 min of ad libitum re-feeding [12].
We hypothesized that food restriction would lead to changes in HPG axis functioning in this species and that some or all of these changes would be reversed by a brief re-feeding period. In order to investigate this hypothesis, we employed three groups of females, ad libitum fed (AL), animals, females food restricted to 60% of ad libitum intake for 48 h (FR), and animals food restricted to 60% of ad libitum for 48 h and then re-fed ad libitum for 90 min (RF). Because others have shown changes in the proportion of GnRH neurons [5] and in GnRH-Fos double-labeling after food restriction [13], we examined proGnRH immunoreactivity (proGnRH-ir) in the entire forebrain and Fos and GnRH double labeling in two regions of the forebrain, the medial septum/diagonal band (MSDB) and the preoptic area (POA). In order to analyze changes in GnRH more quantitatively, we measured GnRH content in tissue punches from the POA and from the median eminence (ME). Finally, we treated females with GnRH and assessed whether feeding condition affected ovulation and corpora lutea (CL) production. We believe that these data add to the growing body of literature on nutritional infertility and extend the knowledge of this field by characterizing food restriction-induced HPG axis deficits in a novel species.
METHODS AND MATERIALS
Subjects and Feeding Procedures
Female musk shrews used for these experiments were born in our colony at the University of Virginia. After weaning, at 21 days of age, females were housed individually in cages (28 by 17 by 12 cm) containing pine wood shavings and paper towels for bedding. The room was kept on a 14L:10D cycle (lights-on at 0600 h, eastern standard time) and only contained adult virgin females. The ambient temperature was 23 ± 1°C. All animals were fed Purina Cat Chow (Ralston-Purina, St. Louis, MO) and provided water ad libitum.
Starting when females were 29–32 days of age and weighed 19–24 g, food intake and body weights were monitored for 4 days. To monitor food intake, the shrews were each given preweighed cat chow (5–6 g) in clean small food cups. All feeding occurred between 1000–1200 h. Twenty-four hours later, the residual food was retrieved and weighed. The average daily intake over the 4-day period was determined for each animal. The females were randomly assigned to one of three groups. The animals in the FR or RF groups were given 60% of their average ad libitum intake for the next 48 h. Females in the AL group continued to be fed as described above. For all experiments, the RF group was given ad libitum access to food 90 min prior to sacrifice or experimental manipulation.
Immunocytochemistry
Twenty to 24 h after the last feeding (0800–1200 h), AL and FR animals were deeply anesthetized with sodium pentobarbital (0.1 mg/kg, i.p.) for perfusion. The RF females were perfused at the same time of day but received ad libitum food 90 min prior to anesthesia. Six AL and seven females from each of the other two groups (FR and RF) were used. Females were perfused first with heparinized saline (100 units heparin/1 ml of 0.9% saline) followed immediately by Zamboni fixative (4% paraformaldehyde in 0.1 M sodium phosphate buffer, pH 7.4, containing 15% saturated picric acid). Brains were removed, cryoprotected overnight in 30% sucrose at 4°C, and quickly frozen using 2-methylbutane in dry ice. Brains were stored at -70°C until they were sectioned.
Brains were sectioned coronally at 30 µm and divided into a series of four wells. One well of tissue (one-quarter of the sections collected) was processed for immunocytochemistry using an antibody to proGnRH (diluted 1:20 000, antiserum #1947) generously provided by Dr. Robert Millar. This antiserum was made against a peptide sequence that spans from the end of the GnRH peptide into the GAP portion (amino acids 6–16), therefore it recognizes only the pro form of mammalian GnRH (mGnRH). A second well of tissue was processed for both Fos and GnRH. The Fos antibody is a sheep anti-Fos antiserum (diluted 1:5000; Cambridge Research Biochemicals, Cheshire, UK). The GnRH antibody (LR-1) is a polyclonal antiserum made in rabbit (used at a 1:50 000 dilution), generously provided by Dr. Robert Benoit. This antibody also recognizes the chicken-II form of GnRH (cGnRH-II), but cGnRH-II cell bodies are not found in the forebrain, so we are confident that the cells we counted were mGnRH [14, 15]. These antisera have been previously validated for use in musk shrew brain [16–18].
Immunoreactivity was visualized using the Vector Elite ABC method (Vector Inc., Burlingame, CA). All rinses and solutions were made in 0.02 M Tris-buffered saline (TBS) (pH 7.8). Sections were pretreated in 1% sodium borohydride to remove residual aldehydes. Tissue was incubated in avidin and biotin blocking solutions (Vector Blocking Kit) to block endogenous biotin. Next, the tissue was incubated in the primary antiserum for 48 h at 4°C. After rinses, tissue was incubated in secondary antiserum (biotinylated goat anti-rabbit IgG for LR-1 and proGnRH and biotinylated rabbit anti-sheep IgG for Fos; Vector, 1:500) for 1 h. After incubation in secondary antiserum, the tissue was treated with avidin-biotin complex. Immunoreactivity was visualized with nickel-intensified diaminobenzidine (DAB) solution (0.25% nickel ammonium sulfate and 0.05% DAB) and activated by 0.001% hydrogen peroxide. For the double-labeling experiment, tissue was incubated first in the Fos primary and taken through the above steps. Next, the sections were rinsed and stored overnight in TBS containing LR-1 antiserum. The development procedure for the LR-1 was the same as above, except nickel ammonium sulfate was not added to the DAB. All tissue was run at the same time for each study to eliminate inter-run variability, and incubation and development times remained constant. Sections processed for Fos were incubated in DAB for 5 min, and the second time through DAB (LR-1) they were processed for 47 min. Sections processed for proGnRH were kept in DAB for 4 min.
Immunocytochemical Data Analysis
The number of GnRH immunoreactive neurons was counted by an observer blind to the treatment of the animals. Labeled neurons were counted in the accessory olfactory bulb (AOB), the nervus terminalis (NT), the MSDB, the medial POA, and the hypothalamus (HT). Gonadotropin releasing hormone cells were visualized using an Olympus BX60 light microscope. Images were projected to a computer screen with an Olympus U-CMAD-2 camera. NIH image was employed to assist with counting and insure that each cell was counted only once. Only GnRH-ir cells that had at least one process connected to the cell body were counted. For the double-labeling experiment, the same criteria were used to identify GnRH neurons. A cell was considered to be double labeled if it met the GnRH neuron criteria and had a darkly labeled nucleus that was clearly in the same plane of section (determined by adjusting the focus and magnification).
Gonadotropin-Releasing Hormone Content
Twenty to 24 h after their final feeding (0800–1200 h), FR and AL (n = 11 per group) females were quickly anesthetized with halothane and decapitated. Females in the RF condition (n = 13) were also sacrificed between 0800–1200 h, but they were given ad libitum food for 90 min before sacrifice. All brains were removed and rapidly frozen in cold 2-methylbutane. Brains were kept frozen at -70° C until sectioning. Brains were sectioned in the coronal plane at 120 µm thickness. Twelve tissue punches were taken from each of two areas (two punches from three consecutive sections of POA and ME for a total of six punches from each area), using an 18-gauge stylus kept on dry ice. The POA punches were collected ventral to the anterior commissure and immediately lateral to the midline (Fig. 1A). The ME tissue was taken from the medial basal HT superior to the infundibulum where the GnRH fiber terminals converge (Fig. 1B).
|
Tissue samples were homogenized in 0.1 N HCl followed by the addition of 3 M NaHCO3 and NaOH to bring the pH to 7.0. The homogenate was centrifuged at 2500 rpm for 10 min. Gonadotropin-releasing hormone content was assessed using a Peptide Enzyme Immunoassay (Peninsula Laboratories, Inc., Belmont, CA). This assay cross-reacts 100% with mammalian GnRH and has a cross-reactivity of 
0.1% with all other forms of GnRH tested. The detection range of this assay is 0–10 ng/ml. All concentrations measured were within this range. This kit was validated for use in musk shrew by running serial dilutions of tissue from the medial basal HT of adult females. The curve generated by this dilution was parallel to the standard curve. All samples were run in duplicate in the same assay (one plate was used for each brain region, and each plate contained a standard curve), to avoid interassay variability. The coefficient of variability averaged 5.38% (POA) and 4.43% (ME). The protein content in each sample was assayed using a Bio-Rad Protein Assay for Microtiter Plates (Bio-Rad Laboratories, Hercules, CA). The linear range of this assay is 0.05–0.5 mg/ml. All GnRH values were corrected for the amount of protein present in each sample and reported as pg GnRH/µg of protein.
Gonadotropin-Releasing Hormone Injection/Ovulation
Musk shrews (n = 71) were given an injection of GnRH (1.25 µg/kg of GnRH in saline, i.p.) between 1200–1400 h. This dose reliably results in ovulation in AL females, and saline treatment never promotes ovulation [19]. The RF animals (n = 25) were treated with GnRH after 90 min of ad libitum feeding. After GnRH treatment, all animals were given their daily amount of food (between 5.0–5.5 g for AL females and between 1.5–2.2 g for FR and RF females). Twenty-four hours later (between 1200–1400 h), females were anesthetized with sodium pentobarbitol and ovariectomized. Ovaries were examined by an observer blind to the feeding condition of the animals. Each ovary was scored for ovulation based on the presence of at least one CL and at least one ovum within the oviduct. The number of CL and ova were recorded. Because ova can be lost, we have less confidence in this measure than our ability to count CL. We therefore report only CL number and not ova number. It should also be noted that the average shrew litter size is two to four pups, so the low number of CL recorded is consistent with the number of offspring usually produced.
Statistics
Differences in the number of GnRH-ir neurons, the number of double-labeled neurons, GnRH content, and the number of CL were analyzed using a one-way ANOVA. Differences among groups were further analyzed using a Student-Newman-Keuls posthoc test. Differences in the percentage of animals that ovulated among the three groups were analyzed using a 3 x 2 chi-squared test. The differences were considered significant if P < 0.05.
RESULTS
Food Restriction Increases proGnRH-ir Neuron Numbers
In the POA and in the POA + HT regions combined, FR females had significantly more proGnRH-ir neurons than shrews in either the RF or AL groups (F2, 17 = 4.09 and 4.10, respectively; P < 0.05). In contrast, the number of proGnRH neurons in all other individual regions (AOB, F2, 17 = 3.38; NT, F2, 17 = 1.87; MSDB, F2, 17 = 0.143; HT, F2, 17 = 2.56), and in the entire forebrain (F2, 17 = 0.668) did not differ by treatment (Table 1).
|
There were very few double-labeled cells in brains taken from females in any of the groups, and no significant differences were observed. The region that had the greatest percentage of double-labeled cells was the MSDB (between 5% and 6.3%), and even here no significant group differences were found (F2, 17 = 0.178).
Food Restriction Increases GnRH Content
There were no differences among the groups in GnRH content in the POA (F2, 32 = 0.844; Fig. 2A). In contrast, the FR females had significantly more GnRH in the ME than either the AL or RF females (F2, 32 = 4.03; P < 0.05; Fig. 2B).
|
Food Restriction Decreases the Number of CL
There were no group differences in the percentage of animals that ovulated in response to the GnRH injection (
2 = 3.05; P > 0.05; Fig. 3A). There were, however, differences in the number of CL present on the ovaries. The FR animals that ovulated (n = 18) had significantly fewer CL than either the RF (n = 21) or the AL (n = 22) females (F2, 68 = 5.34; P < 0.01; Fig. 3B).
|
DISCUSSION
Females food restricted to 60% of ad libitum intake for 48 h displayed two alterations in the GnRH system. First, they had more proGnRH-ir cell bodies in the POA as compared with AL females or those re-fed 90 min prior to sacrifice. Because proGnRH-ir cell numbers were equivalent in AL and RF females, it appears that food reverses this effect on GnRH neurons. These data suggest that GnRH-containing cells can rapidly respond to metabolic fuels. Changes in proGnRH-ir indicate that food intake may also affect the rate at which the GnRH prohormone is processed to its active decapeptide form in specific regions of the forebrain. It is also possible that food restriction increases production of proGnRH. While this explanation seems counterintuitive, we cannot rule out increases in transcription of GnRH mRNA with the semiquantitative techniques used in this study. In future studies we will examine changes in GnRH mRNA after food restriction and re-feeding. We also found very few neurons labeled for both GnRH and Fos in females in any of the feeding conditions. While Fos expression does not always predict gene transcription, it has been shown to increase in GnRH neurons at the time of the LH surge [13], suggesting a functional relationship between Fos and increased GnRH production and/or release. The relative lack of double labeling that we observed in the GnRH neurons suggests that the rapid changes in proGnRH-ir cell number are independent of Fos activation and perhaps independent of GnRH transcription. Others have shown that food deprivation on Days 1 and 2 of the estrous cycle in hamsters significantly reduces the proportion of GnRH neurons that contain Fos protein in the POA and MSDB on Days 2 and 4 of the cycle [13]. If Fos expression is needed in GnRH neurons to facilitate the LH surge during proestrus, it is possible that because shrews do not have cyclic LH surges, the mechanism by which food restriction disrupts the HPG axis does not depend on Fos activation in GnRH neurons. Nappi and Rivest [20] demonstrated that 2 days of food deprivation in gonadally intact, cycling female rats had no significant effect on the number of LHRH neurons containing Fos-ir when compared to AL females. It would appear that in some species, Fos is important for GnRH neuron function, but that it is not critical in other species.
Previous studies have shown that food restriction affects GnRH immunoreactivity and GnRH transcription. Growth-restricted sheep had a larger percentage of GnRH-ir neurons in the medial-basal hypothalamus (MBH) as compared to rapidly grown sheep of the same age, but no differences in the number of GnRH neurons in the POA were observed [5]. The GnRH neurons in the MBH of sheep comprise only 10–20% of the total population of GnRH neurons in the forebrain, while the POA contains 40% to 50% of the GnRH neurons [5]. Thus, while there were changes in the distribution of GnRH neurons after food restriction in this study, the changes were not noted in the primary GnRH neuron population (the POA). We found significant increases in the number of proGnRH-ir neurons in the POA of food-restricted shrews. One difference between our study and I'Anson et al. is that we used an antibody that is specific for proGnRH, while they used the LR-1 antibody that does not distinguish between pro- and GnRH peptide [5, 21]. By using a more specific antibody, we may have identified a level of control of metabolic cues on the GnRH system where reductions in food intake slow down the processing of proGnRH to peptide. Because shrews were food restricted for only 48 h, this reduction in processing may represent the first of several steps along the HPG axis by which limited food resources inhibit reproduction.
The second alteration in GnRH was noted in FR females that had significantly higher GnRH levels in the ME than either the AL or RF females. One interpretation of this finding is that food restriction inhibits GnRH release from the ME. Because AL and RF females had equivalent GnRH content, we hypothesize that re-feeding for 90 min stimulates GnRH release. We would like to conduct an LH assay on plasma from animals treated like these to test the hypothesis that changes in GnRH lead to LH release. Unfortunately, we have not been able to find an LH assay that is sensitive enough to detect LH in musk shrew plasma. We are currently working on developing a reliable LH assay for use in the shrew; those data will no doubt enhance our work greatly. Other studies have shown rapid changes in LH release with re-feeding after food restriction. In fasted male monkeys and in female rats, LH pulses can return within hours after a single meal [22, 23]. In young female rats food restricted to maintain their weights at 80–85 g until 2 mo of age, a return to ad libitum feeding promotes LH surges [6]. This reinstatement of LH pulsatility can occur after a single day of re-feeding. Additionally, when immature female rats are maintained on this schedule of reduced food prior to puberty, growth, and reproductive development are blocked. However, once the animals are returned to unlimited food, ovulation is achieved within 3 days [24]. Based on these rapid changes in LH pulses, it seems likely that metabolic fuels act on the GnRH system in a steroid-independent manner. Despite the differences in GnRH content in the ME, no differences in GnRH content in the POA were found. Although at first, this may appear to contradict our proGnRH-ir data, in fact it is not inconsistent. The enzyme immunoassay used to measure GnRH content utilizes antiserum that recognizes both pro and peptide GnRH. Thus, if food restriction causes a shift in the relative amounts of pro and mature peptide GnRH, the total amount of GnRH would be unchanged. The GnRH content analysis also includes fibers that typically are storage reservoirs for GnRH peptide.
We also found that treatment with endogenous GnRH induced ovulation in close to 100% of the females, regardless of their food condition. Yet, the FR females had fewer CL than the AL or RF females. There are two possible mechanisms that could account for these findings. The first is that food restriction could decrease pituitary sensitivity to GnRH. This is unlikely because several studies have shown that food restriction increases pituitary sensitivity to GnRH [7, 8]. The second possibility is that food restriction retards follicular development in the shrew. There is precedence for this in other species [2]. Printz and Greenwald [25] showed that food deprivation for more than one estrous cycle blocked follicular development in hamsters. In addition, food-restricted sows had lighter ovaries with smaller and fewer follicles when compared to AL sows [26].
Our data describe the effects of food restriction on multiple levels of the HPG axis. These findings demonstrate how environmental conditions impact on fertility. First, we showed that GnRH neurons are able to detect and respond rapidly to changes in food intake. These responses are evident both in the POA and in the ME. Next, we showed that food restriction effects ovarian responsiveness. Our food restriction paradigm might be more similar to conditions in the wild where shrews are probably normally in an underfed state (relative to AL laboratory shrews). The switch from the 60% ad libitum state to the re-fed condition may be a normal mechanism by which the reproductive system responds to a large meal. The reproductive system, particularly in short-lived, opportunistically breeding mammals, likely evolved to be sensitive to transient variations in food availability in order to maximize reproductive efficacy.
ACKNOWLEDGMENTS
We thank Aileen Wills for expert maintenance of our shrew colony. We also thank Julia Perkins, Xia Li, and Patricia Schiml-Webb for technical assistance.
FOOTNOTES
1 This work was supported by RO1 NS 35429 and KO2 MH 01349. J.L.T. was supported by an NSF predoctoral fellowship. 
2 Correspondence. FAX: 804 243 8433; rissman{at}virginia.edu
Accepted: July 20, 2000.
Received: May 1, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  J. S. Schneider and E. F. Rissman Gonadotropin-releasing hormone II: a multi-purpose neuropeptide Integr. Comp. Biol., November 1, 2008; 48(5): 588 - 595. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Kauffman, K. Bojkowska, A. Wills, and E. F. Rissman Gonadotropin-Releasing Hormone-II Messenger Ribonucleic Acid and Protein Content in the Mammalian Brain Are Modulated by Food Intake Endocrinology, November 1, 2006; 147(11): 5069 - 5077. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Temple, R. P. Millar, and E. F. Rissman An Evolutionarily Conserved Form of Gonadotropin-Releasing Hormone Coordinates Energy and Reproductive Behavior Endocrinology, January 1, 2003; 144(1): 13 - 19. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
