|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Departments of Internal Medicine and Cell Biology, University of Virginia, Charlottesville, Virginia 22908
ABSTRACT
GnRH neuronal function is regulated by gonadal hormone feedback. In males, testosterone can act directly or be converted to either dihydrotestosterone (DHT) or estradiol (E2). We examined central steroid feedback by recording firing of green fluorescent protein (GFP)-identified GnRH neurons in brain slices from male mice that were intact, castrated, or castrated and treated with implants containing DHT, E2, or E2+DHT. Castration increased LH levels. DHT or E2 alone partially suppressed LH, whereas E2+DHT reduced LH to intact levels. Despite the inhibitory actions on LH, the combination of E2+DHT increased GnRH neuron activity relative to other treatments, reflected in mean firing rate, amplitude of peaks in firing rate, and area under the curve of firing rate vs. time. Cluster8 was used to identify peaks in firing activity that may be correlated with hormone release. Castration increased the frequency of peaks in firing rate. Treatment with DHT failed to reduce frequency of these peaks. In contrast, treatment with E2 reduced peak frequency to intact levels. The frequency of peaks in firing rate was intermediate in animals treated with E2+DHT, perhaps suggesting the activating effects of this combination partially counteracts the inhibitory actions of E2. These data indicate that E2 mediates central negative feedback in males primarily by affecting the pattern of GnRH neuron activity, and that androgens combined with estrogens have a central activating effect on GnRH neurons. The negative feedback induced by E2+DHT to restore LH to intact levels may mask an excitatory central effect of this combination.
estradiol, gonadotropin-releasing hormone, luteinizing hormone, steroid hormones
INTRODUCTION
GnRH neurons form the final common pathway regulating reproductive function. GnRH is released in pulses [13] that stimulate the synthesis and secretion of the pituitary gonadotropins LH and FSH and are required for fertility in both males and females [47]. Changes in GnRH, and resulting LH, pulse patterns are largely regulated by feedback actions of the steroids released by the ovaries or testes [816]. In males, testosterone is the predominant gonadal steroid in the circulation, but it is often converted in target tissues via aromatization to estradiol (E2) or via 5-alpha reduction to dihydrotestosterone (DHT) [1719].
Although GnRH neurons express the beta isoform of the E2 receptor, many studies indicate that these cells lack the alpha isoform of the E2 receptor and androgen receptors [2024]. This suggests that feedback effects from either steroid are likely indirect, mediated by steroid-sensitive neurons either locally within the hypothalamus or in distant brain regions. In this regard, both androgen and estrogen receptors are found in the hypothalamus of males [21, 2527], and in other areas implicated in steroid feedback regulation [2832]. GnRH neurons and immortalized GnRH neuronal GT1 cells have also been reported to express more recently described membrane-associated forms of both the estrogen and the androgen receptor, suggesting an alternative possibility for direct steroid feedback [33, 34].
There is evidence for both androgen and estrogen action in testosterone-mediated negative feedback. Estrogen action is suggested by suppression of LH by local hypothalamic implants containing E2 but not DHT [35, 36]. In patients lacking estrogen receptors or aromatase activity, LH release is elevated [37, 38]; mouse aromatase knockouts also exhibit elevated LH [39]. Androgen action is suggested by reductions in serum LH levels after systemic treatment with DHT in rodents, sheep, and primates [10, 13, 40]. Additionally, reduction of androgen receptors or blockade of 5-alpha reductase results in elevated LH levels [41, 42]. Reports of feedback actions in estrogen receptor
knock-out mice are mixed, with one study showing an elevation in LH in gonad-intact males [43], suggesting a role for the alpha isoform of the E2 receptor, but others showing no change [35, 44]. In one of the latter studies [44], castration increased, and high-dose DHT treatment reduced LH levels, suggesting androgen receptor-mediated negative feedback. In men with idiopathic hypogonadotropic hypogonadism, testosterone and E2 similarly suppressed LH release from the pituitary in response to a known GnRH signal, suggesting either conversion to E2 or similar action of these two steroids at the pituitary in humans [45, 46]. In contrast, both E2 and dihydrotestosterone were reported to act centrally and not at the pituitary to affect negative feedback on the reproductive neuroendocrine system in sheep [15].
It is thus evident that a consensus has not yet emerged regarding the active steroid(s) mediating testosterone feedback. This may be in part because of species differences, but could also be because of differences in technical approach. Specifically, many studies probe feedback by examining LH levels, which represent the combined central and pituitary actions making positive identification of the site of feedback action difficult. In the present studies we investigated how steroid milieu affects the firing pattern of green fluorescent protein (GFP)-identified GnRH neurons to examine central actions, and compared this to LH levels of male mice.
MATERIALS AND METHODS
Animals
GnRH neurons were recorded from transgenic mice in which GFP is genetically targeted to GnRH neurons [47]. Mice were bred on the premises, housed on a 14L:10D cycle with lights off at 1630 h, and maintained on Harlan 2916 rodent chow (Harlan) and water ad libitum. All procedures were approved by the Animal Care and Use Committee of the University of Virginia.
To examine the effect of steroid milieu on GnRH neurons, adult (>2 mo) male mice were castrated under isoflurane (Abbott Laboratories) anesthesia to remove testis feedback and at that time received steroid implants as described below. We first did a pilot study to determine the dose of DHT that restored seminal vesicle weight to intact levels by treating castrated mice with either one or two Silastic capsules (Dow Corning) containing 400 µg of the nonaromatizable androgen DHT. This pilot study indicated two capsules was the appropriate dose. For the main experiment, intact controls (n = 8 cells in five animals) were compared with castrated (cas, n = 8 cells in seven animals), cas+ E2 (0.625-µg capsule, n = 9 cells in eight animals) cas+DHT (two 400-µg capsules, n = 8 cells in six animals), or cas+E2+DHT (n = 9 cells in six animals). No more than two cells were used from each animal; each cell is considered individually because variation between cells from the same animal was similar to variation within the entire group. All hormones were administered in vivo and were not present in any recording solutions. Postoperative analgesia was provided by a long-acting local anesthetic (bupivacaine, 0.25%, 7.5 µl per site; Abbott Laboratories). Recordings were done 59 days after surgery. There was no difference in any parameter attributable to time after surgery and steroid replacement.
LH Immunoassay and Seminal Vesicle Weight
To confirm endocrine status, LH levels and seminal vesicles were examined. Body and seminal vesicle weight were measured and blood samples were collected at the time of brain slice preparation. LH levels were determined by a previously described modified supersensitive two-site sandwich immunoassay [4850]. The limit of detection (three assays) averaged 0.14 ng/ml; intraassay coefficient of variation (CV) averaged 4.1% and interassay CV averaged 9.9%. E2 was measured by RIA (DSL 39100, Diagnostic Systems Laboratories, Inc.). All samples were run in a single assay with a sensitivity of 1.7 pg/ml and an intraassay CV of 4.3%. Values produced by E2 implants (11.6 ± 1.5 pg/ml, n = 7) in castrates were not different (P > 0.05) from those in intact males (9.5 ± 0.6, n = 3)
Brain Slice Preparation and Recordings
Brain slices were prepared with slight modifications [51] of previous methods [52]. Cell membrane currents associated with action potential firing were recorded using a targeted extracellular recording approach as previously described [53]. The location of each GnRH neuron was mapped on sketches of coronal sections obtained from a mouse brain atlas [54]; no regional differences were observed in any firing parameter among the cells examined. The duration of recordings ranged from 44 to 210 min and was not different between groups (P > 0.10). If no activity was observed for 30 min, 15 mM KCl was added to the bath to check cell viability and recording integrity. If the cell did not respond to KCl, the data set was truncated at the time of last firing. If it fired in response to KCl, the data set was truncated for analysis at the time of adding KCl.
Data Analysis
Using custom programs [52] written for Igor Pro (Wavemetrics), events were counted and binned at 1-min intervals to identify changes in firing properties and at 5-min intervals for Cluster8 analysis to avoid oversampling errors in pulse detection. Cluster8 compares clusters of points by pooled t-testing to look for nadirs and peaks in time series data [55]. Using peak and nadir clusters of one and two points, respectively, peaks were identified and interpeak interval calculated.
Using Excel (Microsoft Corp.), recordings were analyzed for the following parameters: percentage of time in quiescence (
1 event/min), longest duration of quiescence, mean firing rate, instantaneous frequency, area under the curve of frequency, and peak amplitude. Mean firing rate was determined by dividing the total number of events detected by the duration of the recording. Instantaneous frequency is the interval between action potentials converted to frequency. Area under the curve of frequency was determined using Igor Pro and the peak amplitude was calculated by Cluster8. For group comparison, data were log-transformed as needed for normalization and parameters were compared by one-way ANOVA followed by post hoc analysis with Student-Newman-Keuls tests. Significance was set at P < 0.05.
RESULTS
Pilot Study to Determine DHT Dose
Our previous work with DHT [56] established implants that mildly elevate androgen levels in females. To ensure a sufficient male androgen dose, seminal vesicle weight was compared among the treatment groups. Figure 1 shows the expected reduction in seminal vesicle weight with castration. Treatment with one 400-µg DHT capsule did not appear to restore seminal vesicle weight to intact levels (n = 2); thus, two capsules were used in further tests. Treatment of castrate mice with two 400-µg DHT capsules (n = 4 mice, P > 0.1 vs. n = 6 intact mice) restored seminal vesicle weight to intact levels. Although E2 only was not effective in restoring seminal vesicle weight (n = 8 mice, P < 0.05 vs. intact), the combination of E2+DHT elevated seminal vesicle weight slightly above intact levels.
|
LH Levels Reflect Combined Central and Pituitary Steroid Feedback Effects
Serum LH was measured to determine the combined central and pituitary feedback effect of the various treatments (Fig. 2). As expected, castration significantly increased LH levels (P < 0.05). E2 (n = 8 mice) or DHT (n = 6 mice) alone partially suppressed LH levels (P < 0.05 vs. both intact [n = 5 mice] and castrate [n = 6 samples from seven mice; one sample was lost]), whereas LH levels in mice treated with both E2+DHT (n = 6) were not different from intact controls (P > 0.1). These data indicate that both E2 and DHT are required for complete negative feedback at the hypothalamo-pituitary level.
|
Steroid Milieu Affects GnRH Neuron Firing Activity
By examining firing activity of GnRH neurons directly, we can determine how steroid milieu affects individual neurons without the possible confounding effect of pituitary-level feedback actions that alter the response to GnRH. Figure 3 shows examples of raw action current data. A representative action current is shown in Figure 3A. Figure 3, B and C, illustrates 1-min excerpts of raw firing data from two points (indicated by lowercase letters) on the firing rate curves of Figure 4, E2 and A, respectively. The pattern of action of current events is appreciably different between peaks and valleys in firing rate not only when peaks are high-amplitude (Figs. 3B and 4E) but also when peaks are low-amplitude (Figs. 3C and 4A).
|
|
Figure 4 shows representative graphs of GnRH neuron firing activity in the different groups. All data are plotted on the same scale to facilitate comparison. Analysis (mean ± SEM) of various parameters of these data is presented in Figures 5 and 6. Records were evaluated for overall measures of GnRH neuron quiescence, defined as one or more event per minute. Neither percentage of time in quiescence nor maximum duration of quiescence was affected by any treatment (data not shown). Two measures of overall firing activity, mean firing rate and area under the curve of firing rate vs. time, were examined (Fig. 5, A and B). There was no difference in these parameters between intact, castrate, or castrate treated with E2 or DHT alone. In contrast, the combination of E2+DHT induced a marked increase in both of these parameters (P < 0.05). Instantaneous frequency, an indicator of high-frequency burst firing that is associated with more efficient hormone release [57], was increased by castration and by the combination of E2+DHT (both P < 0.05 vs. intact). Treatment with E2 or DHT alone produced values that were not different (P > 0.05) from either castrate or intact values.
|
|
Low-frequency changes in GnRH neuron activity are an important facet of their function; the frequency with which a cell increases firing rate may be correlated with pulses of hormone release. Cluster8 analysis of firing rate data revealed that baseline activity (Cluster8 nadirs) was not affected by any treatment (P > 0.1, not shown). In contrast, castration significantly increased peaks/h (P < 0.05, Fig. 6A) relative to intact controls. E2 alone reduced peak frequency to control levels (P > 0.01 vs. intact). In contrast, DHT alone did not suppress peak frequency (P > 0.1 vs. castrate). Cotreatment with E2+DHT produced a peak frequency that was not different (P > 0.05) from that found in either intact or castrate mice. This may indicate that addition of DHT may partially counteract the suppressive effects of E2 alone. Amplitude of peaks in firing rate was not different among groups, with the exception of a marked elevation in cells of E2+DHT treated mice (P < 0.05, Fig. 6B).
DISCUSSION
Testosterone is the main circulating androgen in males and may be converted to E2 or DHT in target tissues [1719]. The present results suggest that steroid feedback to the reproductive neuroendocrine system in males is both site- and steroid class-dependent. Specifically, E2 provides central negative feedback, whereas both E2 and DHT are required for negative feedback at the pituitary. Direct examination of individual GnRH neurons unmasked this same combination of E2 and DHT as activating at the central level.
A few previous studies have directly examined central actions of steroid feedback in males. Castration increases GnRH release in rats [13] and sheep [58], and testosterone replacement restores GnRH release to intact patterns. Most studies, however, have examined steroid feedback by measuring levels of the pituitary hormone LH [9, 10, 16, 59], which reflects both central and pituitary actions. Further, much previous work used testosterone replacement [10, 12, 14, 16], precluding conclusion with regard to the active steroid.
The present work supports and extends previous studies showing that castration increases LH levels in male mice. Centrally, the main effect of castration on GnRH neuron activity was on the pattern of firing. In this regard, GnRH neurons exhibit patterned firing activity with different frequencies. These include high-frequency, repeated bursts of firing [6062], and low-frequency changes in firing rate [52, 60]. The interval between peaks in firing rate in the latter is similar to that between pulses of GnRH1/LH release [13]. Similar observations have been made examining calcium oscillations in GnRH neurons, with higher-frequency oscillations coordinating to form a lower-frequency rhythm with a period resembling that of GnRH1/LH release [63]. Both low- and high-frequency firing rhythms were affected by removal of gonadal feedback in the present study. Specifically, castration increased the frequency of peaks in firing rate that may be correlated with pulsatile hormone release, and the instantaneous frequency, which is a measure of burst firing. Castration had no effect on more general measures of GnRH neuron activity, such as mean firing rate. This parallels earlier work in ovariectomized and ovariectomized E2-treated females, in which overall activity did not change but the pattern of activity was significantly altered [52].
When castrated mice were treated with steroid replacement, the central and pituitary responses differed. Direct examination of GnRH neuron activity revealed that E2 alone, but not DHT, restored the low-frequency pattern of GnRH neuron activity to that of intact controls. This suggests that the primary negative feedback action of testosterone at the central level is accomplished after its aromatization to E2. Supporting this, E2 implants in the arcuate nucleus/ventromedial hypothalamus reduces LH pulse frequency in castrated sheep and primates [35, 36, 64], whereas localized implants of testosterone or DHT in the same area are ineffective [35, 64]. Further, immunization against E2 [9, 65] or aromatase inhibitors [66] causes increased LH secretion in gonadal intact rams and rats, and humans deficient in the alpha isoform of the E2 receptor or aromatase have elevated or high normal LH levels [37, 38].
Unlike the suppressive action of E2 alone or the lack of action of DHT alone, the combination of these two steroids activated individual GnRH neurons. We had hypothesized that replacing both E2 and DHT would mimic the intact condition. This expected suppressive effect of E2+DHT was observed on LH levels, but not when GnRH neuron activity was examined. These data suggest the possibility that the negative feedback effects at the pituitary mask a central stimulatory effect of combined estrogen and androgen receptor activation that was revealed by direct study of GnRH neurons. Such a masking effect has been observed previously with acute steroid treatment in vivo, but not to our knowledge with chronic feedback such as used in the present study [67, 68]. In considering this hypothesis, it is important to bear in mind two main caveats regarding the approach that was used. First, the activity of individual GnRH neurons may not reflect that of the network as a whole. Second, afferents important for conveying feedback effects to GnRH neurons may have been removed by the slicing procedure. With regard to these caveats, in other studies examining steroid feedback effects on firing activity of individual GnRH neurons, a direct correlation was found between firing activity and reproductive response [52, 69, 70].
At the pituitary, there is evidence that steroids alter response to GnRH, although the effect may depend on species [71]. In rodents in vivo, DHT reduced both serum LH and LH beta gene transcription in response to GnRH1 [48, 7274]. Indeed, the negative feedback effect of DHT on LH levels in mice has been reported to be stronger (i.e., complete suppression) than that observed in the present study [44, 48], likely because of differences in doses. In dissociated rat pituitary cells, testosterone reduced, whereas E2 enhanced, response to GnRH1 [71, 75]. Together with the present data, these observations of direct pituitary negative feedback by androgens support the viability of the hypothesis that pituitary-level steroid negative feedback masks a central activating effect.
The central mechanisms underlying these observations are not known, but there are several possibilities. E2 can induce androgen receptors, perhaps increasing action of DHT [76]. Central synergism between E2 and DHT has previously been reported for the control of aromatase activity in the male rat hypothalamus [77]. Alternatively, in the intact male there may be a near-total conversion of testosterone to E2 in regions regulating GnRH neuron function, so that there is little to no endogenous activation of androgen receptors; E2+DHT treatment would raise the ratio of androgen to estrogen. The high levels of aromatase in male hypothalamus are consistent with this notion [19]. Another possibility is the conversion of DHT to diols that appear to interact specifically with the beta isoform of the estrogen receptor [78]. This is of interest because GnRH neurons have been reported to express mRNA for this isoform [21, 24], although protein has not yet been found in murine GnRH neurons [79]. Further, DHT alone had different effects on GnRH neuron firing activity in the present study than did the combination of E2+DHT.
The potential stimulatory action of E2+DHT in males is of interest with regard to hyperandrogenemic infertility in women. In these disorders, such as polycystic ovary syndrome, androgen-mediated increases in reproductive hypothalamo-pituitary axis activity occur [80]. The present studies suggest that androgens may also activate the central portion of the reproductive neuroendocrine axis in males. Mechanistically, this could be accomplished through changes in synaptic transmission to GnRH neurons, intrinsic properties of these cells, or both. In this regard, DHT elevates gamma-aminobutyric acid (GABA) transmission to GnRH neurons in females [56]. GABAA receptor activation can be excitatory in GnRH neurons [8183], and a preliminary report studying females indicates that firing rate of these cells is also elevated by DHT [70]. Consistent with a possible change in synaptic transmission, DHT, but not E2, increases dendritic spine density in CA1 hippocampal pyramidal neurons in male rats [84].
Together, the present results indicate that E2 is the main hormone that provides negative feedback at the central level, whereas the pituitary requires both E2, acting centrally or directly, and DHT for a complete negative feedback effect. Future studies will explore the synaptic and intrinsic mechanisms underlying these effects, as well as the activating effects of these steroids in combination, to understand at which levels the feedback systems of males and females differ.
ACKNOWLEDGMENTS
We thank Catherine Christian, Zhiguo Chu, Debra Fisher, Alison Roland, and Pei-San Tsai for editorial comments.
FOOTNOTES
1 Supported by the National Institute of Child Health and Human Development/National Institutes of Health through cooperative agreement U54HD28934 and the Ligand Assay and Analysis Core. A preliminary report of this work was presented at the 38th annual meeting of the Society for the Study of Reproduction, 2427 July 2005, Quebec City, Quebec, Canada. ![]()
2 Correspondence: Suzanne M. Moenter, Department of Internal Medicine, P.O. Box 800578, University of Virginia, Charlottesville, VA 22908. FAX: 434 9820088; smm4n{at}virginia.edu ![]()
Received: 23 November 2005.
First decision: 8 December 2005.
Accepted: 31 January 2006.
REFERENCES
This article has been cited by other articles:
![]() |
Z. Chu and S. M. Moenter Physiologic Regulation of a Tetrodotoxin-Sensitive Sodium Influx That Mediates a Slow Afterdepolarization Potential in Gonadotropin-Releasing Hormone Neurons: Possible Implications for the Central Regulation of Fertility. J. Neurosci., November 15, 2006; 26(46): 11961 - 11973. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |