| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
Departments of Pediatrics (V.P.), Physiology (F.J.K.), and Biostatistics (M.B.B., D.T.M.) and the Reproductive Sciences Program (V.P., G.E.D., N.P.E., F.J.K., J.V.C.), The University of Michigan, Ann Arbor, Michigan 48109; and Departments of Physiology and Biophysics (J.D.N.), University of Alabama-Birmingham, Birmingham, Alabama 35294
Address all correspondence and requests for reprints to: Vasantha Padmanabhan, University of Michigan, Reproductive Sciences Program, 300 North Ingalls Building, Room 1109 SW, Ann Arbor, Michigan 48109-0404. E-mail: vasantha{at}umich.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Using an approach to access the hypophyseal portal system (8, 9), we have been able to decipher FSH secretory dynamics more accurately (10). Because the hypophyseal portal vessels at the surface of the pituitary are lesioned for collection of portal blood, this approach provides an avenue to monitor the secretory dynamics of pituitary hormones close to the site of their release (11). The utility of this approach has helped demonstrate unequivocally that an episodic mode of FSH secretion exists (10). These studies also documented the existence of a dual mode of FSH secretion (basal and episodic). Furthermore, these studies clearly documented that each GnRH pulse is associated with an FSH pulse and that additional pulses of FSH not associated with GnRH/LH pulses are present. In those studies, however, it was not possible to prove whether these additional FSH pulses were acute responses to GnRH pulses that were below the threshold of GnRH assay sensitivity, or whether they were triggered by a separate stimulus. Neuroanatomical and biochemical evidence exists in support of a separate FSH-releasing factor (12, 13, 14), although none has been isolated so far. The current studies were undertaken to test the hypothesis that a GnRH-independent component of episodic FSH secretion exists. The approach we took to address this question was to monitor FSH secretory patterns near the site of release in ovariectomized and luteal phase ewes after blockade of GnRH action with Nal-Glu, a competitive GnRH receptor antagonist.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Is there a GnRH-independent component of episodic FSH secretion?
To determine whether a GnRH-independent component of episodic FSH secretion exists, we studied two animal models: 1) an ovariectomized model characterized by high-frequency GnRH pulses and elevated circulating LH and FSH concentrations (8, 16, 17); and 2) a luteal model characterized by low-frequency GnRH pulses and low LH and FSH concentrations (17). Studies in ovariectomized sheep (n = 5; BW, 68–89 kg) were conducted during the anestrous season. Studies in luteal ewes (n = 6; BW, 50–85 kg) were conducted during the breeding season on d 7–10 after estrus. Estrous behavior was monitored with the aid of raddled vasectomized rams. Luteal status of the ewes was confirmed at autopsy by the presence of corpora lutea on ovaries and by plasma progesterone measurements.
All experiments were conducted using adult Suffolk ewes that were maintained outdoors under natural photoperiod at the Sheep Research Facility (Ann Arbor, MI; latitude, 42° 18' N). Animals were fed hay and provided free access to water and mineral blocks. Details of the preparation of animals for the collection of hypothalamo-pituitary portal blood have been described previously (8, 9). Briefly, ewes were surgically fitted with an apparatus for collection of hypophyseal portal blood (in the luteal model, d 1–4 after detection of estrus), and the collection procedure was initiated approximately 1 wk later. Integrated samples of hypophyseal portal and peripheral (will be referred to hereafter as jugular, reflective of the site from which blood was collected) blood were collected at 5-min intervals for 12 h in tubes containing 0.5 ml of 3 x 10-3 M bacitracin in phosphate-buffered isotonic saline. Midway through the collection (6 h), animals were given a bolus injection of Nal-Glu (10 µg/kg BW iv). The collection process is described fully in Ref. 9 . For hypophyseal portal collection, the pump speed for collection of hypophyseal portal blood was set at a higher speed than the rate of flow of portal blood into the collection apparatus. This facilitates collection of hypophyseal portal blood as discrete blocks segmented by air and minimizes dispersion of secreted products during collection. Because portal vessels are cut and not cannulated, hypophyseal portal blood collects by passive flow into a small well (about 50 µl) of the collection apparatus placed just anterior to the pituitary. As the well fills, portal blood is removed from the top of the pooled blood in the well. The physics of the collection system does not facilitate preferential suction of FSH-laden blood up into the collection system, thus creating artificial FSH pulses. Furthermore, because both LH and FSH are secreted by the same gonadotrophs, changes in flow dynamics will be reflected in both LH and FSH. As such, the collection approach is suited for identifying non-GnRH/LH-associated pulses of FSH.
At the end of the period of portal blood collection, all ewes were euthanized with a barbiturate overdose (Beuthanasia, Schering-Plough Animal Health Corp., Kenilworth, NJ), and the site and extent of the lesion were examined to ensure that the surgical approach to the pituitary and placement of the collection apparatus had been successful. The University of Michigan Committee on the Use and Care of Animals approved all experimental procedures.
Does Nal-Glu block interaction of GnRH II with GnRH type II receptors?
In the above studies, levels of GnRH in the hypophyseal portal samples were measured using GnRH I (mammalian GnRH [pGlu His Trp Ser Tyr Gly Leu Arg Pro Gly NH2]) as the standard. The competitive GnRH receptor antagonist Nal-Glu used in this study to eliminate GnRH action blocks the interaction of GnRH I with GnRH type I receptor. Provided Nal-Glu does not block the interaction, generation of the non-GnRH-associated pulses of FSH may be facilitated via interaction of the newly identified second mammalian GnRH, GnRH II (18, 19), with its receptor, GnRH type II receptor (20, 21, 22). To determine whether Nal-Glu distinguishes between interaction of GnRH with GnRH I and II receptors, we experimentally expressed the two receptors in separate sets of COS-1 cells and measured the inhibitory effects of Nal-Glu on 3H-inositol phosphate production stimulated by GnRH I and GnRH II (chicken GnRH II; p Glu His Trp Ser His Gly Trp Tyr Pro Gly NH2). COS-1 cells were cultured at 37 C in DMEM supplemented with 10% fetal bovine serum as previously described (23). GnRH receptors were expressed in the COS-1 cells using cDNA-mediated transfections (23); 16.7 µl of Lipofectamine (Life Technologies, Inc., Gaithersburg, MD) was mixed with 1.6 µg pcDNA 3.1 carrying either the human GnRH I receptor cDNA (24) or the porcine GnRH II receptor cDNA (25). The resulting Lipofectamine-DNA complexes were incubated with COS-1 cells for 5–8 h before DMEM/20% fetal bovine serum was added (23). The transfected cells were used in inositol phosphate accumulation assays at 48 h after the beginning of transfection.
To measure GnRH-stimulated total inositol phosphate accumulation, we labeled cells in 60-mm dishes with 4.0 µCi [3H]myoinositol for 16 h, beginning 24 h after transfection as previously described (23). Then the cells were incubated for 1 h in the presence of 5 mM LiCl with or without GnRH I or GnRH II. The cells were then disrupted and centrifuged, and the supernatant fluids were used to isolate [3H]inositol phosphates using Dowex 1x8–400 (Dow Chemical Co., Midland, MI) ion exchange resin (23). Eluted [3H] samples were measured using liquid scintillation spectroscopy.
RIAs
Concentrations of LH in hypophyseal portal and jugular samples were assayed in duplicate (26). The jugular samples were assayed in 200 µl aliquots, and hypophyseal portal samples were assayed in aliquots of 5–20 µl. The LH assay sensitivity (2 SD values from the buffer control) averaged 0.16 ± 0.03 ng/tube (n = 20 assays). The intraassay coefficient of variation at 80 and 20% displacement points averaged 6.3 ± 0.3 and 3.1 ± 0.1%, respectively. The median variance ratio averaged 0.039 ± 0.003. The interassay variability based on two quality control pools with 2.54 and 26.3 ng/ml LH·ml averaged 6.8 and 9.3%, respectively. Plasma FSH concentrations in hypophyseal portal and jugular samples from both ovariectomized and luteal phase ewes were measured using reagents distributed by the National Hormone and Pituitary Program (NIH antibody). All jugular or hypophyseal portal samples from a given animal were measured in a single assay using 10- or 80-µl aliquots of hypophyseal portal and jugular plasma, respectively. The assay sensitivity and the median variance ratios of this assay averaged 27 ± 8 pg/tube and 0.018 ± 0.002, respectively (n = 14 assays). The intraassay coefficient of variation at 80 and 20% displacement points averaged 6.1 ± 0.4 and 3.0 ± 0.2%. The interassay variability based on three quality control pools containing 4.3, 11.0, and 22.2 ng/ml FSH averaged 15.6, 14.7, and 12.0%, respectively.
Because FSH is secreted as a mix of isoforms (2, 3), to assess whether the antibody used in the NIH FSH assay failed to recognize some of the FSH isoforms, thus masking the true secretory pattern of FSH, hypophyseal portal and jugular samples from the ovariectomized animals were also measured using a well validated second RIA that uses antibody no. 620 (27, 28). For measurement of FSH using 620 antibody, 5-µl aliquots of hypophyseal portal plasma and 40-µl aliquots of jugular plasma were used. National Institute of Diabetes and Digestive and Kidney Diseases ovine FSH-3 was also used as the standard in this assay. The assay sensitivity of this second RIA averaged 10 ± 6 pg/tube (n = 6 assays). The intraassay coefficient of variation at 80 and 20% displacement points averaged 11.8 ± 0.4 and 5.9 ± 0.2%, respectively. The median variance ratios averaged 0.032 ± 0.005, and interassay variability based on three quality control pools containing 4.0, 17.8, and 231.6 ng/ml FSH·ml averaged 16, 18, and 17.0%, respectively.
Statistical analysis
To allow direct comparisons with concentrations of LH and FSH in jugular samples (nanograms per milliliter) but at the same time to adjust for subtle changes in flow, all measurements of FSH as well as LH in hypophyseal portal plasma were first calculated as collection rate and then multiplied by 5 to convert to concentrations in portal blood (referred to as concentrations in the Results section and tables). Hypophyseal portal measurements from each ovariectomized and luteal ewe were analyzed using Kushler-Brown pulse detection algorithms (29). The Kushler-Brown algorithm is based on a nonlinear statistical model incorporating exponential decay between distinct pulses. The model allows for both assay error and biological variability in the observed time series. Pulses identified by the algorithm were screened using the criteria that differences between peak and previous nadir must exceed assay sensitivity. To be conservative in our estimates, hormone series were also subjected to pulse analysis after smoothing each series using a three-point running average. Although marginal differences in pulse frequency estimates were evident, with unsmoothed series providing a higher estimate of pulse frequency, conclusions drawn were qualitatively similar. Results from the smoothed series are presented.
Hormone data from ovariectomized animals were also analyzed with the Cluster algorithm (30). The Cluster algorithm identifies pulses using a criterion that defines a pulse such that the peak of the pulse differs significantly from both the preceding and following nadirs according to two-sample t tests. For analysis with Cluster, the minimum number of data points in a peak and nadir were set at one and two, respectively. The t statistic values used to identify a significant increase from preceding nadir and a decrease to following nadir were both 3.8. Pulses are identified using a stepwise nonlinear regression procedure. Because the results from Cluster analysis were qualitatively similar to that of Kushler and Brown, only results from Kushler-Brown analysis, which was performed on both ovariectomized and luteal series (Cluster was performed only on ovariectomized series), are reported in the Results section.
After pulse analysis, concordance of pulses of GnRH and hypophyseal portal LH, GnRH and jugular LH, and GnRH and hypophyseal portal FSH were identified. Due to the difficulty encountered in identifying clearly definable pulses of FSH, jugular FSH series were not subjected to pulse analysis. For concordance estimates before and after Nal-Glu, pulses beginning within 10 min of each other were considered to be concordant. Pulses of GnRH that started at the very end of the pre-Nal-Glu period but continued into the post-Nal-Glu period and the associated LH/FSH pulses that appeared in the post-Nal-Glu period were excluded from analysis. Percentage concordance for each ewe was calculated as a proportion (x 100%) of total pulses. This value was compared with the percentage of concordant pulses that would be expected if pulses were occurring randomly. The expected percentage of random concordance was estimated by calculating the theoretical expectation of concordance under the assumption that the locations of the pulses in each series were randomly distributed, with at least 5 min of quiescence between adjacent pulses. Paired t tests were used to test the effect of Nal-Glu on pulse frequency and pulse amplitude and to compare observed and expected concordances. In addition, to determine temporal relationships between hormone series, auto cross correlation was performed, and cross correlations for GnRH, LH, and FSH were calculated using different time lags. An estimate of the overall time lag between two series of hormones (GnRH/LH and GnRH/FSH) was obtained from the time lag yielding the highest cross correlation.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
depicts the patterns of GnRH, hypophyseal portal LH (P-LH), and jugular LH (J-LH) before and after administration of a GnRH antagonist, Nal-Glu. Patterns of GnRH and J-LH have been reported earlier (31) and are presented for comparison of relationships with hypophyseal LH patterns. As reported earlier, GnRH pulse frequency increased post Nal-Glu in three of the five animals studied (31). Comparison of LH patterns in hypophyseal portal and jugular blood exemplifies the validity of this approach in using hypophyseal portal blood as a tool for monitoring pituitary hormone secretory dynamics. As one would expect, LH patterns in both the hypophyseal portal and jugular blood showed a distinctively episodic pattern of release. Concentrations of P-LH were severalfold higher than J-LH, and the patterns of LH were more discrete in the hypophyseal portal than jugular blood. No LH pulses were evident in both the hypophyseal portal and jugular circulation after administration of Nal-Glu, although measurable basal secretion of LH was still present. There was a one-to-one concordance between GnRH and LH during the pre-Nal-Glu period but not during the post-Nal-Glu period due to the absence of LH pulses.
|
depicts unsmoothed hypophyseal portal FSH (P-FSH) patterns as measured by both the NIH and 620 antibodies in the same ovariectomized ewe and the jugular FSH (J-FSH) as measured by the NIH antibody. In contrast to the distinctly episodic nature of J-LH, patterns of J-FSH measured with the NIH antibody were relatively stable, with little evidence of episodic secretion (Fig. 2, bottom). J-FSH patterns as measured by the 620 antibody were similar to that of NIH (data not shown). J-FSH patterns contrasted markedly with the patterns of P-FSH, as measured by two different antibodies (NIH and 620). An episodic pattern of P-FSH secretion was clearly evident in both instances (Fig. 2, top and middle panels). Before Nal-Glu, as reported earlier for ovariectomized ewes (10), P-FSH secretion was distinctly pulsatile, with an FSH pulse evident after most GnRH pulses in the hypophyseal portal circulation (asterisks indicate pulses identified after smoothing of data). Importantly, in addition to these GnRH-associated pulses, additional episodes of FSH release were evident (Fig. 2, arrows). Some of the pulses, which met the criteria of non-GnRH-associated pulses of FSH, also showed a secondary increase in FSH corresponding to the GnRH pulses suggestive of overlapping triggers (e.g. first two FSH pulses). After Nal-Glu administration, the discreteness of GnRH-associated pulses of FSH disappeared, although an episodic pattern of FSH secretion continued to persist. Pulses of FSH during the post-Nal-Glu period appeared to be of much smaller amplitude than that of the GnRH-associated pulses during the pre-Nal-Glu period. The pattern of FSH identified by the two FSH antibodies in the ovariectomized animals was qualitatively similar, although absolute levels of FSH measured were higher with the 620 antibody (Fig. 2, middle). Figure 3 depicts comparative GnRH, hypophyseal portal LH, and FSH from another ovariectomized ewe, after smoothing of data, documenting similar relationships.
|
|
depicts the GnRH and hypophyseal portal and jugular patterns of LH and FSH in a luteal phase ewe before and after administration of Nal-Glu. GnRH pulse frequency increased post Nal-Glu, suggesting the existence of a short or ultra-short loop feedback at the level of GnRH secretion (31). As in the ovariectomized ewes, LH patterns in the hypophyseal portal blood were severalfold higher than those in the jugular circulation and more discrete. Before Nal-Glu administration, P-LH pulse accompanied each GnRH pulse. No LH pulses were evident after administration of Nal-Glu, although basal release of LH still persisted.
|
depicts comparative GnRH, P-LH, and P-FSH from another luteal phase ewe, documenting similar relationships.
|
summarizes the mean number of hypophyseal portal and jugular LH and FSH pulses during the pre- and post-Nal-Glu periods. Although results from unsmoothed series were qualitatively similar to smoothed series, to be conservative only results from smoothed series are discussed. Before Nal-Glu treatment, the pulse frequency of P-LH was not different from that of GnRH in both the ovariectomized and luteal groups, confirming their one-to-one relationship. In both ovariectomized and luteal models, the pulse frequency of P-FSH was far greater than that of GnRH or LH. After Nal-Glu, in both the ovariectomized and luteal groups, LH pulsatility was completely obliterated, but P-FSH pulses persisted. There were no statistically discernable differences between FSH pulse frequency before and after Nal-Glu treatment in both the ovariectomized and luteal groups, consistent with overlapping triggers (e.g. first two pulses in Fig. 2
).
|
summarizes the pulse concordance relationships of GnRH, P-LH, and P-FSH (NIH antibody) in the ovariectomized and luteal phase ewes before and after Nal-Glu treatment. Results with the 620 antibody were qualitatively similar to those derived with NIH antibody and are not provided. In the ovariectomized model, during the pre-Nal-Glu period, a GnRH pulse accompanied every P-LH pulse, resulting in a 100% concordance between GnRH and P-LH pulses (P < 0.001). No pulses of LH existed in the absence of a GnRH pulse. The same relationship holds true when data were analyzed using the Cluster algorithm (data not shown). In the luteal model, 97% of the GnRH pulses were associated with hypophyseal portal and jugular LH pulses (P < 0.01). During the post-Nal-Glu period in both the ovariectomized and luteal models, no P-LH or J-LH pulses existed.
|
Time lag relationships between GnRH, LH, and FSH pulses
Autocorrelation analysis revealed that GnRH pulses precede P-LH pulses by 0–5 min (after correcting for collection time lag) in both the ovariectomized and luteal model during the pre-Nal-Glu period and that this relationship disappears after administration of Nal-Glu. The time lag relationship was similar for GnRH and P-FSH, with GnRH leading GnRH-associated FSH pulses by 0–5 min and this relationship disappearing after Nal-Glu administration.
LH and FSH pulse amplitudes
Table 3 summarizes the mean (±SE) amplitudes of the GnRH, P-LH, and P-FSH pulses during the pre- and post-Nal-Glu period. To gain a true perspective of the magnitude of pulse amplitude, results from unsmoothed series are provided. Amplitudes of GnRH-associated pulses of P-FSH during the pre-Nal-Glu period were higher than the non-GnRH-associated pulses of P-FSH during the same period as well as amplitudes of P-FSH pulses during the post-Nal-Glu period (P < 0.05 in ovariectomized animals; did not reach statistical significance in luteal animals). There were no differences between the GnRH-associated and non-GnRH-associated FSH pulse amplitudes during the post-Nal-Glu period of ovariectomized and luteal animals, consistent with the association being random (data not shown). In general, pulses of FSH measured by the 620 antibody were 2.5-fold higher than the FSH pulses measured by the NIH antibody.
|
. As reported earlier, total GnRH measured during the post-Nal-Glu period was significantly higher than during the pre-Nal-Glu period in the luteal group but not the ovariectomized group (31). In both models, the difference in levels between jugular and hypophyseal portal levels were higher for LH (25- and 63-fold for ovariectomized and luteal groups, respectively) compared with FSH (11- and 19-fold for ovariectomized and luteal groups, respectively). In both the ovariectomized and luteal models, total P-LH measured and total J-LH in circulation were significantly higher in the pre- than the post-Nal-Glu period (P < 0.01). During the post-Nal-Glu period, although the pulsatile component of P-LH was obliterated, LH continued to be secreted even after blockade of GnRH action. In contrast to LH, total FSH measured in hypophyseal portal blood did not differ between the pre- and post-Nal-Glu periods in both the ovariectomized and luteal groups, suggesting that the amount of FSH secreted in the acute GnRH-associated pulses of FSH is negligible in relation to the total amount of FSH secreted.
|
, A and B. The EC50 dose of GnRH I to stimulate the GnRH I receptor was approximately 1 x 10-10 M (Fig. 6A). The minimal and maximal inhibitory effects of Nal-Glu were approximately 1 x 10-9 M and 1 x 10-8 M, respectively, with the IC50 dose of approximately 5 x 10-9 M (Fig. 6A).
|
). The EC50 of GnRH II (
1 x 10-9 M) was not inhibited even at 10-6 M Nal-Glu (Fig. 6B). Used alone, Nal-Glu had no agonistic activity (data not shown). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Existence of a constitutive component of LH release
In a previous study, we showed that termination of the active phase of secretion during an LH pulse was followed by continued release of LH during the interpulse interval (10, 11). The present study extends these initial observations and provides evidence for the existence of a constitutive component of GnRH-independent basal LH release, even after Nal-Glu blockade of GnRH action. The discreteness of the LH pulses, the one-to-one relationship of GnRH with LH and their immediate blockade after GnRH antagonist administration during the pre-Nal-Glu period, and the similarity in levels of interpulse LH during the pre-Nal-Glu period and the constant persisting levels of LH secretion during the post-Nal-Glu period support incremental active secretion and not leakage from damaged pituitary cells. Damaged cells are likely to deplete their content and not maintain a constant secretion.
Documentation of acute GnRH-independent episodic component of FSH secretion
Earlier studies that monitored FSH at the peripheral level after blockade of GnRH action with GnRH antiserum plus GnRH antagonist (4) or after blocking generation of GnRH pulses with the aid of 
-adrenergic antagonist phentolamine (32) predicted a GnRH-independent component of episodic FSH secretion. The present study, using hypophyseal portal blood as a resource to access pituitary secretions, provides the first direct and definitive evidence of the existence of an episodic component of FSH secretion that is not acutely regulated by GnRH I. The episodic pattern of FSH secretion continued after the blockade of GnRH I action with a competitive GnRH antagonist in two different endocrine states, the luteal phase and following ovariectomy.
Measures taken to determine true dynamics of FSH secretory patterns
Before evaluating the clues that the FSH secretory patterns during Nal-Glu blockade provide toward our understanding of FSH control, it was essential to assess whether changes in FSH secretion are evaluated accurately. Because FSH is secreted as a mix of FSH isoforms (2, 3), it is critical that the assay used, although specific for FSH, is not discriminatory in terms of recognizing the various FSH isoforms. With this in mind, we verified the patterns of FSH in ovariectomized ewes obtained with the use of the polyclonal NIH antibody using a second assay that used a different polyclonal antibody (antibody 620). Both assays have been used in the past, are specific for FSH, exhibit very little cross reactivity with LH or gonadotropin , and recognize the intact, deglycosylated, and asialo variants of ovine pituitary FSH with equal potency (33, 34). Patterns of FSH obtained with both assays are qualitatively similar, suggesting that the two antibodies recognize the various FSH isoforms secreted. The clearly definable episodic patterns of FSH in hypophyseal portal but not peripheral circulation suggest that the episodic component of FSH secretion may be comprised of faster-clearing FSH isoforms. The massive release at the secretory site may have allowed clear recognition of the fast-clearing forms before they have had a chance to clear from the circulation. Findings in GnRH-deficient subjects, which show that pulsatile administration of GnRH leads to pulsatile release of fast-clearing FSH isoforms that are recognizable in a bioassay but not immunoassay (35), corroborate this premise. Our recent studies in sheep with desialylated FSH also document that the faster clearing isoforms of FSH are harder to estimate in the periphery (34). The absolute differences in levels of FSH measured using two specific and well validated FSH assays, despite the use of the same FSH standard, suggest that the two different antibodies recognize FSH with different affinity or, alternatively, that the NIH antibody is not detecting one of the FSH isoforms that is secreted in a similar manner. Within the context of the question posed in this study, both assays yielded qualitatively similar answers, namely documentation of the existence of a component of episodic FSH secretion that was independent of an acute GnRH I drive and suggestive of other neural or paracrine mechanisms.
Possible means by which the acute GnRH I-independent episodic component of FSH secretion is generated
There are several possible mechanisms by which the episodic FSH secretion seen after Nal-Glu treatment can be explained. We have reviewed these possibilities in detail in a recent review (14). These possibilities will be discussed briefly in the context of the results of this study. The first possibility is that persistence of episodic FSH secretion following Nal-Glu is the outcome of incomplete blockade of classical GnRH receptor (type I) to which Nal-Glu binds. Complete blockade of LH pulses, the biomarker of GnRH release, but not FSH pulses following Nal-Glu administration; the existence of only LH pulses but not FSH pulses in association with the low magnitude GnRH pulses in luteal-phase ewes (this study); and available evidence that supports a higher threshold of GnRH requirement for releasing FSH than LH (15) argue against this possibility. The observed GnRH/FSH concordance during the post-Nal-Glu period in the ovariectomized ewes appears to be the result of chance concordance because of the high-frequency GnRH/gonadotropin pulses that prevail in this model. This is also borne out by the fact that the observed GnRH/FSH concordance between GnRH/FSH during the post-Nal-Glu period is not statistically different from random concordance estimates of GnRH/FSH.
The second possibility is that there is an intrinsic pituitary FSH rhythmicity. This is not likely because FSH secretion from perifused ovine pituitary cells and ovariectomized animals during halothane anesthesia, a central nervous system depressant, are nonpulsatile (Padmanabhan, V., N. P. Evans, C. Herkimer, and J. Lee, unpublished observations).
A third possibility is that the non-GnRH-associated episodic pattern of FSH secretion may be the outcome of activation by GnRH of a second class of GnRH receptors that are unaffected by Nal-Glu, the proven competitive GnRH antagonist of the well studied mammalian GnRH receptor (type I). Several studies have shown the existence of an additional form of GnRH receptor (type II) in higher vertebrates (primates, Refs. 20 and 21 ; humans, Ref. 22) or differently sized GnRH receptor transcripts (ovine; Ref. 36). Of interest, the type II receptor, which is expressed in gonadotropes, has a C-terminal tail and desensitizes (21, 22). For these receptors to be involved in the selective regulation of non-GnRH-associated pulses of FSH and satisfactorily explain the selective presence of FSH pulses during the post-Nal-Glu period, two lines of evidence are required. It is essential to show first that Nal-Glu is ineffective in blocking signal transduction through GnRH type II receptor, and second that GnRH or a variant form of GnRH is capable of releasing FSH in the absence of LH secretion. It is unlikely that GnRH I stimulated the GnRH II receptor in these studies because GnRH I has 1/430 the potency of GnRH II at the GnRH II receptor (22). For this to happen, GnRH I levels in stalk blood would need to be about 430 ng/ml (0.43 µg/ml). Given that GnRH I levels in hypophyseal portal blood are at most less than 1 ng/ml, this possibility does not appear likely. The interaction of GnRH II with the GnRH I receptor was also precluded in our studies because the GnRH I receptor was blocked by Nal-Glu. On the contrary, our studies document that Nal-Glu does not inhibit GnRH II signal transduction through GnRH type II receptor. Considering that GnRH II has been reported to stimulate preferential secretion of FSH (21), GnRH II and its receptor cannot be ruled out as playing a regulatory role in the generation of non-GnRH-associated pulses of FSH.
The fourth possibility is that the non-GnRH-associated episodes of FSH secretion are the outcome of temporally dissociated changes in endocrine/pituitary paracrine milieu triggered by the GnRH stimulus. Key candidates for such endocrine or local paracrine action are inhibin and activin, members of the TGFß family, and follistatin, a binding neutralizer of activin (for review, see Refs. 14 and 37, 38, 39). These regulators modulate FSH production and consequently affect basal FSH release (37, 39). Inhibin and follistatin inhibit FSH, and activin stimulates FSH production (37). Existing evidence suggests that GnRH regulates the production of activin/inhibin subunit, follistatin mRNA (for review, see Refs. 40 and 41), and inhibin release (42). A possibility to consider is that episodic release of GnRH leads to episodic release of activin, inhibin, and/or follistatin, which in turn leads to increases or decreases in FSH secretion. A delay in this sequence would provide a basis for the presence of non-GnRH-associated pulses of FSH secretion during the post-Nal-Glu period. In the present study, we examined FSH pulsatility only up to 6 h post Nal-Glu. Earlier studies report a delay of 6–15 h between inhibin/follistatin treatment and FSH response (43, 44, 45). Our preliminary studies showing obliteration of FSH pulses during a 6-h halothane anesthesia (Padmanabhan, V., N. P. Evans, C. Herkimer, and J. Lee, unpublished observations) argue against mediation via changes in paracrine milieu.
A fifth and a likely possibility is that the non-GnRH-associated pulses of FSH are responses to a yet to be identified hypothalamic FSH-releasing factor (FSH-RF). There is neuroanatomical, biochemical, and physiological evidence to support the existence of a selective FSH-RF. This has been reviewed extensively in the literature (12, 13, 14). The most convincing anatomical evidence comes from studies that demonstrate ablation of FSH pulsatility without corresponding alteration in LH pulsatility after radiofrequency lesions of the dorsal anterior hypothalamic area (46) or lesions of the caudal and midmedian eminence (47). Biochemical studies also support this contention and have shown FSH-releasing activity to be localized in the posterior median eminence, whereas GnRH activity is confined predominantly to the anterior median eminence (47). Although a putative FSH-RF that appears to differ from GnRH has been isolated (48), the recent demonstration of variant forms of GnRH (18, 19, 49), difficulties in achieving a clear separation of GnRH and FSH-RF activity (12, 13, 48), and the close proximity of FSH-RF and GnRH activity during molecular sieving (48) suggest that a GnRH variant may be the long-sought FSH-RF. Recent studies in rats suggest that lamprey GnRH-III selectively stimulates FSH at lower doses (50). However, Lamprey GnRH-III was equally efficacious in stimulating LH and FSH secretion from cultured ovine pituitary cells (Lee, J. S., S. M. McCann, and V. Padmanabhan, unpublished observations). Evidence supporting the selective release of FSH by this variant form of GnRH in the presence of Nal-Glu is also lacking. Preliminary studies using ovine pituitary cultures do support the existence of an acute FSH-releasing factor in the hypophyseal portal circulation of the ewe (Lee, J. S., F. J. Karsch, V. Padmanabhan, unpublished observations).
Relevance of the low-amplitude non-GnRH-associated pulses of FSH
Considering that both the GnRH-associated and non-GnRH-associated pulses of FSH are considerably dampened in the periphery, it is essential that we address the relevance of the episodic component of FSH in the context of target site function. One possibility is that the non-GnRH-associated FSH pulses may have no physiological relevance in distant organs such as the ovary. Alternatively, pulsatile delivery of fast-clearing FSH isoforms could still induce significant, albeit relatively brief but repetitive, biological responses. In support of this concept, prior studies have documented that 1) the pulsatile component of FSH (2, 3, 35) is enriched with fast-clearing isoforms of FSH, which provide a basis for their dampening in the periphery; 2) GnRH-induced increases in FSH are more clearly evident when measured using a bioassay than immunoassay (35); and 3) rapid functional interactions are possible between FSH and the FSH receptor. For instance, calcium increases are reported to occur within minutes after addition of FSH to target cells (51, 52, 53). Effective activation of adenylyl cyclase, under conditions of intermittent receptor activation, also appears possible because activated G proteins can have a half-life in the order of seconds (54). Further support for this concept is provided by our recent studies in which administration of deglycosylated FSH, despite its fast clearance, yielded significant biologic outcomes (34). Another possibility that remains to be explored is that the episodic component of FSH is involved in setting the local activin/inhibin tone, thus indirectly influencing the basal component of FSH release, the predominant mode of FSH secretion (10).
In summary, the present study, using Nal-Glu to effectively block GnRH action, provides the first conclusive evidence for the existence of an episodic component of FSH secretion that is independent of acute GnRH I stimulation. The likely trigger for which most evidence is available points to acute mediation by a selective hypothalamic trigger or, alternatively, a selective time-lagged response to GnRH II that is mediated through GnRH receptor type II.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Present address for N.P.E.: Department of Veterinary Preclinical Studies, University of Glasgow Veterinary School, Bearsden Road, Glasgow, G61 1QH, United Kingdom.
Present address for D.T.M.: Center for Biostatistics and Epidemiology, The Pennsylvania State University College of Medicine, 500 University Drive, Hershey, Pennsylvania 17033.
This work was supported by U.S. Public Health Service Grants HD-34731, U54-HD-29184, and NSF IBN 9725943.
Abbreviations: BW, Body weight; FSH-RF, FSH releasing factor; J-FSH, jugular FSH; J-LH, jugular LH; P-FSH, portal FSH; P-LH, portal LH.
Received September 16, 2002.
Accepted for publication January 6, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. N. Sarma, M. Manikkam, C. Herkimer, J. Dell'Orco, K. B. Welch, D. L. Foster, and V. Padmanabhan Fetal Programming: Excess Prenatal Testosterone Reduces Postnatal Luteinizing Hormone, But Not Follicle-Stimulating Hormone Responsiveness, to Estradiol Negative Feedback in the Female Endocrinology, October 1, 2005; 146(10): 4281 - 4291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Amstalden, D A Zieba, M R Garcia, R L Stanko, T H Welsh Jr, W H Hansel, and G L Williams Evidence that lamprey GnRH-III does not release FSH selectively in cattle Reproduction, January 1, 2004; 127(1): 35 - 43. [Abstract] [Full Text] [PDF]
|
|
| ||||
) or with an unvarying dose of GnRH I (10-9 M; 
) in addition to varying doses of Nal-Glu (10-11 to 10-7 M). The EC50 of GnRH I alone was approximately 10-10 M, and the IC50 of Nal-Glu was about 10-9 M. B, GnRH II receptor. Aliquots of cells were treated with varying doses of GnRH II (10-11 to 10-7 M;