Downloaded from

Transcription

1 BIOLOGY OF REPRODUCTION (2012) 86(6):171, 1 9 Published online before print 14 March DOI /biolreprod Neuroendocrine Control of FSH Secretion: IV. Hypothalamic Control of Pituitary FSH-Regulatory Proteins and Their Relationship to Changes in FSH Synthesis and Secretion 1 Tejinder P. Sharma, 3,4 Terry M. Nett, 5 Fred J. Karsch, 4 David J. Phillips, 6 James S. Lee, 3 Carol Herkimer, 3 and Vasantha Padmanabhan 2,3,4 3 Department of Pediatrics, University of Michigan, Ann Arbor, Michigan 4 Reproductive Sciences Program, University of Michigan, Ann Arbor, Michigan 5 Animal Reproduction and Biotechnology Laboratory, Colorado State University, Fort Collins, Colorado 6 Monash Institute of Medical Research, Monash University, Clayton, Victoria, Australia ABSTRACT The current dogma is that the differential regulation of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) synthesis and secretion is modulated by gonadotropinreleasing hormone (GnRH) pulse frequency and by changes in inhibins, activins, and follistatins both at the pituitary and at the peripheral level. To date no studies have looked at the overlapping function of these regulators in a combined setting. We tested the hypothesis that changes in GnRH pulse frequency alter the relative abundance of these regulators at the pituitary and peripheral levels in a manner consistent with changes in pituitary and circulating concentrations of FSH; that is, an increase in FSH will be accompanied by increased stimulatory input (activin) and/or reduced follistatin and inhibin. Ovariectomized ewes were subjected to a combination hypothalamic pituitary disconnection (HPD)-hypophyseal portal blood collection procedure. Hypophyseal portal and jugular blood samples were collected for a 6-h period from non-hpd ewes, HPD ewes, or HPD ewes administered GnRH hourly or every 3 h for 4 days. In the absence of endogenous hypothalamic and ovarian hormones that regulate gonadotropin secretion, 3-hourly pulses of GnRH increased pituitary content of FSH more than hourly GnRH, although these differences were not evident in the peripheral circulation. The results failed to support the hypothesis in that the preferential increase of pituitary content of FSH by the lower GnRH pulse frequency could be explained by changes in the pituitary content of inhibin A, follistatin, or activin B. Perhaps the effects of GnRH pulse frequency on FSH is due to changes in the balance of free versus bound amounts of these FSH regulatory proteins or to the involvement of other regulators not monitored in this study. gonadotropin-releasing hormone, gonadotropins, neuroendocrinology, pituitary, pituitary hormones, sheep INTRODUCTION It is well established that gonadotropin-releasing hormone (GnRH) regulates the release of both luteinizing hormone LH 1 Supported by National Institutes of Health grant HD to V.P. 2 Correspondence: Vasantha Padmanabhan, Reproductive Sciences Program, 300 N. Ingalls Bldg., Rm NW, Ann Arbor, MI vasantha@umich.edu Received: 13 December First decision: 4 January Accepted: 27 February Ó 2012 by the Society for the Study of Reproduction, Inc. eissn: ISSN: and follicle-stimulating hormone (FSH) [1]. In spite of this common mediator, divergent circulating patterns of LH and FSH are often encountered both in physiological and in pathophysiological situations. Earlier studies of Wildt and Knobil [2] provided the first evidence that changes in patterns of GnRH input to the pituitary are important for differential patterns of LH and FSH secretion, with lower frequencies favoring FSH and higher frequencies LH. The influence of GnRH pulse frequency in facilitating divergent LH/FSH release appears to involve changes in pituitary paracrine/ autocrine regulators of FSH, namely, activins, inhibins, and follistatins [3 6]. Activins and inhibins, members of the transforming growth factor b family of growth and differentiation factors, and follistatins, binding proteins that neutralize activins, are major regulators of FSH production [7, 8]. Activins are dimers of either b-subunits (activin A, ba-ba; activin B, bb-bb; activin AB, ba-bb), while inhibins are dimers of a structurally dissimilar a- and b-subunits (inhibin A, a-ba; inhibin B, a-bb) [7, 9]. Inhibins can antagonize the action of activins [10 13]. It is believed that the relevant form of activin that regulates FSH secretion is activin B, as neutralizing antiserum directed against the activin bb-subunit but not ba-subunit profoundly suppresses FSH secretion [14, 15]. Follistatins exist as a number of isomers, with the most characterized known as follistatin 288 and follistatin 315 [8]. Follistatins can bind both activins and inhibins through their b-subunits [16]. Activins maintain activity only when not bound to follistatins, and the binding of follistatins to inhibin has a lower affinity [17]. In addressing the role of these FSH-regulatory proteins in mediating FSH secretion, it is important to consider the relative presence of the various regulators from both an endocrine (peripheral) and paracrine (intrapituitary) perspective [18, 19]. While large strides have been made in support of a role for inhibins, activins, and follistatins in the control of FSH, we are far from gaining a full understanding of the nature of this complex regulation. For example, are the differential effects of GnRH pulse frequency on FSH release due to an altered equilibrium of inhibins, activins, and follistatins? Namely, is there a shift toward higher activin bioavailability (stimulatory input to FSH) when FSH is high and reduced activin bioavailability (reduced stimulatory input to FSH) when FSH is low? Several in vitro studies have documented changes in steady-state levels of pituitary follistatins and subunit mrnas of inhibins and activins as a function of changes in GnRH pulse patterns and consistent with changes in levels of expression of FSHB mrna [3 6]. For example, exposure of perifused rat 1 Article 171

2 SHARMA ET AL. pituitary cells to different frequencies of GnRH results in an inverse relationship between mrna expression patterns of FSHB and follistatins [3]. Similar studies relating changes in FSH to intrapituitary measures of these regulators at the protein level are limited. One study using pooled pituitary extract found an inverse relationship between pituitary concentration of follistatin and FSH during the rat estrous cycle [20]. A limitation in our current understanding of FSH regulation is that there is no study examining all of these regulatory proteins in unison in spite of their overlapping functions. We tested the hypothesis that changes in GnRH pulse frequency differentially regulate LH and FSH and alter the relative abundance of FSH regulators at the pituitary and peripheral levels in a manner consistent with changes in pituitary and circulating concentrations of FSH, that is, an increase in FSH being accompanied by increased activin and/or decreased inhibin and follistatin, and vice versa. MATERIALS AND METHODS Animal Model Sheep were used as the experimental animal because secretory patterns of GnRH can readily be monitored in pituitary portal blood. To address the influence of changes in GnRH pulse frequency in modulating the relative equilibrium of FSH-regulatory proteins, in the absence of other hypothalamic factors that regulate the anterior pituitary, we utilized a combination hypothalamic pituitary disconnection (HPD)-hypophyseal portal collection approach. This approach allows comparison of the profile of delivered pattern of GnRH with the endogenous GnRH pulse profile near the site of delivery to the pituitary. In general, absence of LH pulses following HPD is used as an index of successful elimination of endogenous GnRH input to the pituitary. Ovariectomized sheep were used in the study to assess the effect of GnRH pulse frequency in the absence of ovarian steroids and inhibins. Experimental Design Twenty-three Suffolk ewes ovariectomized (OVX) during the nonbreeding season (June July) of two successive years were utilized to study the effect of GnRH pulse frequency in the absence of other hypothalamic and ovarian factors. A month after ovariectomy, 18 of the 23 OVX ewes underwent surgical HPD using a modification [21] of a transnasal-transphenoidal procedure developed by Clarke et al. [22]. A piece of Teflon was placed between the hypothalamus and the pituitary to provide a physical barrier that ensures that hormonal communication between the hypothalamus and pituitary is eliminated. At the time of HPD surgery, all ewes were simultaneously fitted with an apparatus for the collection of hypophyseal portal blood using a previously described approach [23, 24]. Five non-hpd controls were also fitted with this hypophyseal portal collection apparatus. Seven days after surgery, blood samples were obtained from all HPD animals at 10-min intervals for 5 h to determine if LH pulsatility was ablated. LH pulses were evident in two of the 18 animals, suggesting HPD was incomplete; these ewes were removed from the experiment. The remaining 16 animals were assigned to one of three groups. Group 1 served as an HPD control (n ¼ 6) and received no further treatment. Group 2 received hourly injections of GnRH (n ¼ 5), and group 3 received injections of GnRH every 3 h (n ¼ 5). These frequencies are within the range observed in ovary-intact ewes during the follicular and luteal phases of estrous cycle, respectively [25]. GnRH (2 ng/kg body weight) was administered manually as i.v. injections beginning 7 days after HPD surgery. All sheep were heparinized, and the pituitary portal vessels were lesioned to initiate hypophyseal portal blood flow as described earlier [24]. After establishment of stable blood flow, hypophyseal portal (for measurement of GnRH and secretory patterns of gonadotropins) and peripheral (for determination of peripheral patterns of gonadotropins) blood samples were collected at 5-min intervals for 6 h beginning 90 h after start of GnRH administration (last 6 h of 4-day GnRH treatment). GnRH measurement in hypophyseal portal circulation served two purposes: 1) to validate that hypothalamic input to the anterior pituitary had been eliminated by HPD and 2) to provide a reference for establishing if the amount and pattern of exogenous GnRH delivered to the pituitary was physiological. Hypophyseal portal samples were collected into tubes containing 0.5 ml of M bacitracin (Sigma Chemical Co., St. Louis, MO) in PBS, using an automated sampling procedure as previously described [26]. This collection approach defines the temporal nature of GnRH signal reaching the pituitary. At the end of collection, all ewes were euthanized with a barbiturate overdose (Beuthanasia, Schering-Plough Animal Health Corp., Kenilworth, NJ) to confirm appropriate lesion of pituitary vasculature. Visual inspection showed no infarction of the anterior pituitary. Pituitaries were weighed and snap frozen in liquid nitrogen for the subsequent determination of pituitary content of LH, FSH, inhibin A, activin A, activin B, and total follistatins. Each pituitary was later homogenized in radioimmunoassay buffer and centrifuged at g, and supernatants were stored at 708C as stock concentrations of 0.5 mg pituitary/ml for the assessment of pituitary hormonal content. All experimental procedures were approved by the University of Michigan Committee on the Use and Care of Animals. Hormone Assays Hypophyseal portal samples were extracted in methanol and duplicate aliquots of the extract assayed for GnRH using a well characterized radioimmunoassay [26]. The GnRH assay sensitivity (2 SD from the buffer control) and ED 50 were pg and pg (mean 6 SE, n ¼ 6 assays), respectively. The intra-assay coefficient of variation (CV) was %, and the interassay CV, based on a quality control pool averaging pg/ml, was 10.9%. LH concentrations in hypophyseal portal and jugular plasma were determined using a validated competitive double antibody radioimmunoassay [27]. Measurement of gonadotropins in hypophyseal portal blood provides a better assessment of secretory dynamics [28]. The assay sensitivity (2 SD from the buffer control) and 50% displacement point of the LH assay was ng ( ng/ml, n ¼ 7 assays) and ng/tube, respectively. The mean intra-assay CVs at 80% and 20% displacement points were % and %, respectively. The interassay CVs in three quality control pools averaging , , and ng/ml were 21.8%, 6.2%, and 5.7%, respectively. FSH was measured in duplicate samples (n ¼ 6 assays) with a validated radioimmunoassay [29] using reagents from the National Hormone and Pituitary Program. The assay sensitivity (2 SD from the buffer control) and 50% displacement point of FSH assay were ng ( ng/ ml, n ¼ 6 assays) and ng/tube, respectively. The mean intra-assay CVs at 80% and 20% displacement points were % and %, respectively. The interassay CVs based on two quality control pools averaging and ng/ml were 7.9% and 8.0%, respectively. Total activin A (bound plus free) was measured in pooled 2-hourly plasma samples using a specific ELISA (Oxford Bio-Innovations, Oxfordshire, U.K.) validated for use in sheep [30, 31]. The standard used was human recombinant activin A. Standards and samples were diluted in 5% BSA/0.01M PBS. Crossreactivity of inhibin A, inhibin B, and activin A in this assay is reported to be less than 0.5% [31]. A solution of 6% sodium dodecyl sulfate was added (3% final concentration) followed by boiling for 3 min to liberate activin A from any binding proteins. Activin A concentrations in pituitary homogenates were measured after minor modifications to the original protocol: standard and samples were diluted in 0.05% BSA in PBS to match the protein concentration of the samples. The samples were allowed to cool before the addition of H 2 O 2 (2% final concentration) and subsequent 30-min incubation. Twenty-five microliters of 5% BSA/0.1 M Tris/5% Triton X-100/0.9% NaCl/0.1% NaN 3 were added to each well prior to the addition of the treated samples. Duplicates were added to the anti-ba subunit monoclonal antibody-coated plate and incubated overnight at room temperature. The plates were washed, and the second detection antibody (biotinylated anti-ba subunit monoclonal antibody) was added for 2 h at room temperature. After washing, alkaline phosphatase linked to streptavidin was added to the wells and incubated at room temperature for 1 h. After further washes, the alkaline phosphatase activity was detected using an amplification kit (ELISA Amplification System; Invitrogen). Crossreactivity of inhibin A and activin B in this assay is reported to be less than 0.5% [31]. The limit of detection for the activin assays was 0.01 ng/ml. For the plasma samples, the average intra-assay CV was 6.7%, and the interassay CV was 7.2% (n ¼ 3 assays). For the pituitary extracts, the intra-assay CV was 3.8%, and the interassay CV was 4.1% (n ¼ 2 assays). Total activin B (bound plus free) in pituitary extracts and pooled peripheral plasma samples were measured using a recently developed ELISA [32], which uses monoclonal antibody 46A/F as both the capture and the detection antibody. Cross-reactivity of inhibin A, inhibin B, and activin A in this assay is reported to be less than 0.1% [32]. Activin B standards were prepared in PBS containing 5% BSA, using recombinant activin B from R&D Systems Europe Ltd (Abingdon, Oxfordshire, U.K.). Validity of the activin B assay for sheep samples was confirmed by showing that increasing volumes of ovine pituitary extracts as well as peripheral plasma yielded potency estimates in parallel with the standard curve (Fig. 1). The assay sensitivity, intra-, and interassay CVs averaged ng/ml, 3.5% and 9.2%, respectively. Immunoreactive inhibin in peripheral samples and pituitary extracts was measured by a heterologous radioimmunoassay [33] validated for use in sheep, 2 Article 171

3 HYPOTHALAMIC CONTROL OF FSH-REGULATORY PROTEINS FIG. 1. Activin B assay validation. Shown are parallelism curves of activin B standard, pituitary extracts (P1, P2, P3, and P4), and plasma (J1, J2, and J3) obtained from ovariectomized ewes. using human recombinant inhibin as standard and tracer. The assay cross-reacts 288% with pro-ac, the prosequence of the inhibin a-subunit. For serum/plasma samples, castrate ram serum, which contained no detectable inhibin A, was used as diluent to eliminate any matrix effect of the plasma. For pituitary extracts, the diluent was PBS with BSA matched to the protein concentration of the samples. For the plasma samples, the average intra-assay %CV was 7.9%, and the interassay %CV was 2.2% (n ¼ 2 assays). For the pituitary extracts, the average intra-assay %CV was 9.2%, and the interassay %CV was 10.5% (n ¼ 3 plates). The limit of detection for the all assays was 0.21 ng/ml. Total inhibin A concentrations were measured using an ELISA specific for ovine inhibin [34] using anti-ba subunit monoclonal coated plates (Oxford Bio-Innovations, Oxfordshire, U.K.) and modified using a biotinylated monoclonal antibody (PPG-1) specific for the ovine a-subunit [35]. Crossreactivity of inhibin B, activin A, and activin B in this assay is reported to be less than 0.1% [34]. The standard used was a pool of bovine follicular fluid calibrated in this assay against the WHO 91/684 inhibin A standard. For plasma samples, castrate ram plasma, which contained no detectable inhibin A, was used as diluent to eliminate any matrix effect of the plasma. For pituitary extracts, the sample diluent was PBS with BSA matched to the protein concentration of the samples. For the plasma samples, the average intra-assay CV was 3.5%, and the interplate CV was 7.5% (n ¼ 6 plates). For the pituitary extracts, the average intraplate CV was 7.1%, and the interplate CV was 6.6% (n ¼ 2 plates). The limit of detection for the all assays was ng/ml. Measures of inhibin B were not undertaken, as earlier studies found no detectable inhibin B levels in ram circulation or ovine follicular fluid [36]. Total follistatin concentrations were measured using a discontinuous radioimmunoassay as described previously [37]. This detects total follistatin using a reagent that dissociates the activin-follistatin complex. For pituitary extracts, the diluent was PBS with BSA matched to the protein concentration of the samples. For the plasma samples, the average intra-assay CV was 7.8%, and the interassay CV was 3.1% (n ¼ 2 assays). The limit of detection was 2.30 ng/ ml. For the pituitary extracts, the average intra-assay CV was 11.7%, and the interassay CV was 4.5% (n ¼ 3 assays). The limit of detection was 1.60 ng/ml. Statistical Analysis Data from one HPD control ewe and one HPD ewe receiving 3-hourly injections of GnRH were incomplete and excluded from analyses since sample collection was terminated because of low hematocrit. This resulted in a sample size of five ewes for the non-hpd control, HPD control, and HPD hourly GnRH groups and four ewes for the HPD 3-hourly GnRH group. GnRH and LH pulses (both at the hypophyseal portal and at the peripheral level) were identified using the Cluster algorithm [38] for assessment of frequency, total GnRH/LH secreted within a pulse (area), and pulse duration. Only pulses with upslopes and downslopes were included in estimates of pulse frequency and amplitude. FSH data were not subjected to pulse analyses because of the predominant constitutive pattern of release. The Cluster algorithm identifies pulses using criteria that defines a pulse such that the peak of the pulse differs significantly from both the preceding and the following nadirs according to two-sample t-tests. For LH analysis with Cluster, the minimum number of data FIG. 2. Representative patterns of GnRH and jugular LH and FSH from each of four treatment groups (non-hpd control, HPD control, HPD ewes given hourly GnRH, and HPD ewes given 3-hourly GnRH). GnRH was administered for 96 h, and samples were collected during the last 6 h (91 96 h). points in a peak and nadir were set at 2 and 2, respectively, and the t-statistic values used to identify a significant increase from preceding nadir and a decrease to following nadir were also set at 2.0. For GnRH, the minimum number of data points in a peak and nadir were set at 1.0 and 1.0, respectively. Differences in all measured variables for GnRH, LH, FSH, and the FSH regulatory proteins between the four treatment groups were compared by analysis of variance after appropriate transformations (square root for number of pulses and log transformation for all other variables) to test for changes in mean GnRH, LH, FSH, activin A, activin B, inhibin A, follistatin concentrations, GnRH and LH pulse frequency, GnRH and LH pulse area, and duration of GnRH pulse. All statistical analyses were carried out with the aid of the Statview for Macintosh statistical computer package. Post hoc testing was by Scheffe t-statistic. Significance was defined as P, RESULTS Patterns of Hypophyseal Portal GnRH and Peripheral LH and FSH Representative patterns of GnRH in hypophyseal portal blood and LH and FSH in peripheral blood from a non-hpd control, HPD control, and HPD ewes treated with hourly and 3- hourly boluses of GnRH are presented in Figure 2. As expected, high-frequency endogenous GnRH and LH pulses were found in non-hpd control ewes, with each GnRH pulse accompanied by an LH pulse. There were no GnRH or LH pulses in HPD control ewes. Administration of hourly or 3- hourly boluses of GnRH resulted in GnRH and LH patterns consistent with the delivery pattern of GnRH. As has been previously reported, clearly distinguishable episodes of FSH release were not evident in the peripheral circulation of any of the groups. GnRH Pulse Characteristics Figure 3 summarizes the GnRH pulse characteristics in non- HPD control, HPD control, and HPD ewes receiving exogenous GnRH boluses. GnRH pulse frequency in non- HPD controls averaged approximately five pulses during the 6- h collection period (Fig. 3A). The GnRH frequency in HPD 3 Article 171

4 SHARMA ET AL. FIG. 3. GnRH dynamics. A) Frequency estimates derived from measures of GnRH in hypophyseal portal blood. Differing letters indicate significant differences (P, 0.05). B) Patterns of endogenously produced (non-hpd controls) and exogenously delivered (composite from HPD ewes administered hourly and 3-hourly GnRH) patterns of GnRH. C) Mean 6 SEM of GnRH pulse amplitude (Amp), area, and duration (bottom). Asterisks indicate significant differences (P, 0.05). animals receiving hourly or 3-hourly GnRH boluses was consistent with what was expected from the delivery pattern. The GnRH frequency in HPD animals receiving hourly GnRH resembled the endogenous GnRH pulse frequency of the non- HPD controls (Fig. 3A). The pattern of GnRH measured in the hypophyseal portal circulation produced by exogenous GnRH boluses, however, differed from that of endogenous GnRH pulses (Fig. 3B). Specifically, the amplitude, total amount, and duration of exogenous GnRH were all greater (P, 0.05) than corresponding values for endogenous GnRH pulses from non- HPD ewes (Fig. 3C). LH-FSH Secretory Characteristics Figure 4 summarizes LH/FSH secretory characteristics determined from measures of LH and FSH in hypophyseal portal blood for the four groups. LH pulse frequency in hypophyseal portal blood mirrored GnRH frequency and showed 100% concordance with GnRH in all groups (Fig. 4A, left panel). Total LH secreted per pulse in non-hpd controls and HPD ewes receiving hourly GnRH pulses were similar (Fig. 4A, right panel). Total LH secreted per pulse of sheep receiving 3-hourly pulses of GnRH was significantly greater than those receiving hourly pulses of GnRH or the non- HPD ewes (P, 0.05) (Fig. 4A, right panel). FSH was detectable in all groups, including HPD controls not receiving GnRH. Since clearly definable episodes of FSH release were FIG. 4. A) Mean LH pulse frequency, total LH secreted in each LH pulse during the 6-h sampling period from non-hpd controls, HPD controls, and HPD ewes given hourly or 3-hourly pulses of GnRH. B) The variances associated with FSH measured during the same period. These estimates are based on measures of LH and FSH in hypophyseal portal circulation. For comparison, GnRH frequency estimates from Figure 3 are coplotted as gray bars behind LH frequency estimates. Differing letters indicate significant differences (P, 0.05). not evident for FSH, variance across sampling window was used as an index for changes in secretory dynamics. Variance associated with FSH across the sample series was low (P, 0.05) in the HPD controls compared with HPD ewes receiving 3-hourly administration of GnRH or non-hpd controls (Fig. 4B). Figure 5A summarizes mean peripheral LH and FSH concentrations over the 6-h collection period and LH/FSH ratio in the four treatment groups. Peripheral LH was reduced.90% and near detection limit in HPD controls not treated with GnRH compared with non-hpd controls. Mean LH in those animals receiving hourly administration of GnRH for 4 days did not differ from that of the non-hpd controls. Threehourly GnRH for 4 days also increased mean LH concentrations relative to HPD controls but only to 50% of non-hpd controls. Mean FSH levels 11 days post-hpd were reduced by ;80% in HPD animals not receiving GnRH compared with non-hpd controls. Both hourly and 3-hourly boluses of GnRH for 4 days restored FSH to values seen in non-hpd controls. LH/FSH ratio of HPD animals receiving hourly pulses of GnRH did not differ from non-hpd or HPD controls. LH/FSH ratio in those receiving 3-hourly pulses of GnRH was lower than the non-hpd controls and HPD ewes receiving hourly pulses of GnRH. LH/FSH ratio of HPD controls not treated with GnRH was lower than that in the non-hpd controls. LH/ FSH ratio of HPD ewes receiving 3-hourly GnRH differed from those receiving hourly GnRH and non-hpd controls. GnRH Frequency Modulation of Pituitary LH and FSH Figure 5B summarizes pituitary content of LH and FSH in the four groups of animals. Pituitary LH content was similar across groups (left panel). Pituitary FSH content was higher in HPD ewes receiving 3-hourly boluses of GnRH for 4 days compared with HPD controls receiving no GnRH or hourly GnRH for 4 days (right panel) but similar to that of non-hpd controls. Pituitary weights did not differ across treatment groups (data not shown). 4 Article 171

5 HYPOTHALAMIC CONTROL OF FSH-REGULATORY PROTEINS GnRH Frequency Modulation of Circulating FSH Regulatory Proteins Measurable levels of immunoreactive inhibin were found in circulation and did not differ between treatment groups (data not shown). Dimeric inhibin A levels were undetectable in all four groups. Plasma activin A levels were higher in HPD animals compared with non-hpd controls (Fig. 6A). Plasma activin A levels were not altered by hourly or 3-hourly boluses of GnRH compared with HPD controls. Plasma activin B levels tended to be higher in HPD controls compared with non-hpd controls (Fig. 6B). Plasma activin B levels were higher in HPD animals receiving GnRH at hourly or 3-hourly intervals compared with non-hpd controls. Plasma levels of follistatin were higher in HPD animals receiving no GnRH compared with non-hpd animals (Fig. 6C). Three-hourly but not hourly boluses of GnRH for 4 days restored circulating follistatin levels back with non-hpd control levels. FIG. 5. A) Mean circulating levels of LH, FSH, and LH/FSH ratio from non-hpd control, HPD control, HPD ewes given hourly GnRH, and HPD ewes given 3-hourly GnRH groups. B) Pituitary LH and FSH content from same groups of animals. Differing letters indicate significant differences (P, 0.05). GnRH Frequency Modulation of Pituitary Content of FSH Regulatory Proteins Immunoreactive inhibin levels were reduced in HPD controls compared with non-hpd controls (Fig. 7A). Immunoreactive inhibin levels of HPD animals receiving 3-hourly administration of GnRH were similar to non-hpd controls but differed significantly from the HPD controls. Immunoreactive inhibin levels of HPD animals receiving hourly administration of GnRH were intermediate between HPD controls and HPD ewes receiving 3-hourly pulses of GnRH but did not differ significantly from the other three groups. There were no differences in dimeric inhibin A levels across treatment groups (Fig. 7B). Activin A was not detectable in pituitary homogenates. Pituitary activin B levels did not differ between treatment groups (Fig. 7C). Follistatin levels were similar in non-hpd controls, HPD controls, and ewes receiving hourly injections of GnRH (Fig. 7D). Three-hourly boluses of GnRH increased follistatin levels significantly compared to the other three groups. DISCUSSION Our findings provide supportive evidence that, in the absence of hypothalamic and ovarian input, 3-hourly pulses FIG. 6. Circulating concentrations of activin A (A), activin B (B), and total follistatin (C) in the non-hpd control, HPD control, HPD ewes given hourly GnRH and HPD ewes given 3-hourly GnRH groups. Differing letters indicate significant differences (P, 0.05). 5 Article 171

6 SHARMA ET AL. TABLE 1. Changes in direction of gonadotropins and FSH regulatory proteins in HPD controls relative to non-hpd controls. a Hormone Peripheral Pituitary LH ## NC FSH ## NC Immuno-inhibin NC # Inhibin A UD NC Activin A " UD Activin B "* NC Follistatin " NC a NC, no change; UD, undetectable. * P ¼ FIG. 7. Pituitary concentrations of immunoreactive inhibin (A), dimeric inhibin A (B), activin B (C), and total follistatin (D) in the non-hpd control, HPD control, HPD ewes given hourly GnRH, and HPD ewes given 3-hourly GnRH groups. Differing letters indicate significant differences (P, 0.05). of exogenous GnRH increase pituitary content of FSH more than hourly GnRH and that these differences in pituitary FSH content were not reflected in peripheral FSH concentrations. However, the relative balance of pituitary inhibin A, follistatin, and activin B in response to GnRH pulse frequency were not consistent with what would be predicted from expected changes in FSH production and release. These findings as they relate to the current dogma of GnRH frequency modulation of LH and FSH release and the role of pituitary paracrine regulators of FSH are discussed below. GnRH Frequency Modulation of Gonadotropins Current dogma is that changes in GnRH pulse frequency differentially regulate LH and FSH release with low-frequency GnRH selectively increasing FSH release, although evidence exists to the contrary [39 48]. In the present study, the hourly and 3-hourly frequencies of GnRH resulted in similar magnitude increase in both FSH and LH release in ovariectomized sheep devoid of endogenous GnRH stimulation of the pituitary, although a selective increase in FSH content at the pituitary level was evident with the lower-frequency GnRH stimulation. Earlier studies with gonadectomized HPD sheep are inconsistent. Reducing GnRH pulse frequency from high (one pulse per hour) to low (one pulse every 2 or 4 hours) preferentially increased FSH release [41]. On the contrary, increasing GnRH pulse frequency from low to high frequency (4 h to hourly) did not reduce FSH release [41]. Studies with gonad-intact rams found increasing GnRH pulse frequency from low to high (2-hourly to hourly) reduced FSH release, while the converse, namely, going from high to low (hourly to 2-hourly), had no effect on FSH secretion [42]. These divergent outcomes relative to frequency modulation of FSH secretion in the previous studies and our study may stem from 1) sequential testing of GnRH frequency in the same animals in earlier studies as opposed to use of separate animals in this study, 2) breed differences (Merino in earlier study vs. Suffolk in this study), 3) dose of GnRH (500 ng in the earlier studies vs. 150 ng in this study), or 4) differences in duration of GnRH treatment prior to monitoring LH and FSH (1 wk in earlier study vs. 4 days in our study). Considering that lowerfrequency GnRH pulses for 4 days increased FSH within the pituitary but not in peripheral circulation, it is conceivable that, with longer treatment, the increased pituitary FSH content seen in the 3-hourly GnRH treatment group might be translated into preferential FSH increases in peripheral circulation. GnRH Frequency Modulation of Regulators of FSH Peripheral level. The absence of dimeric inhibin A in the circulation in this study is consistent with the premise that the ovary is the primary source of circulating inhibin A [49] and the fact that ewes in this study were ovariectomized. Measurable levels of immunoreactive inhibin in circulation likely represent inhibin a subunit from various sites since this assay detects both dimeric inhibin and inhibin a subunit [33]. While evidence points to likely presence of inhibin B at several sites including pituitary, brain, and the adrenal in various species [50, 51], inhibin B has not been detected in circulation or ovine follicular fluid [36]. The directional changes in circulating levels of regulatory proteins in HPD control (increased levels of activin A, activin B [trend], and total follistatin) relative to non-hpd controls (Table 1) suggests that there is an inhibitory influence of hypothalamic regulators on circulating pools of activins and total follistatin. Hypothalamic and/or pituitary regulators of activins and follistatin distinct from GnRH that might be altered following HPD have not been well defined. Potential candidates include those that stimulate follistatin mrna, such as thyrotropin-releasing hormone [52] and pituitary adenylate cyclase-activating polypeptide [53] and 6 Article 171

7 HYPOTHALAMIC CONTROL OF FSH-REGULATORY PROTEINS TABLE 2. Changes in direction of gonadotropins and FSH regulatory proteins in HPD ewes receiving GnRH versus HPD or non-hpd control. a GnRH versus HPD control GnRH versus non-hpd control Circulating Pituitary Circulating Pituitary Hormone Hourly 3-Hourly Hourly 3-Hourly Hourly 3-Hourly Hourly 3-Hourly LH "" " NC NC NC NC # NC FSH "" "" NC " NC NC NC NC Im-inhibin NC NC NC " NC NC NC NC Inhibin A UD UD NC NC UD UD NC NC Activin A NC NC UD UD NC NC UD UD Activin B NC NC NC NC " " NC NC Follistatin NC NC NC " " NC NC " a NC, no change; UD, undetectable. cortisol, which has been shown to inhibit activin b-subunit mrna and protein [54]. From an FSH regulatory perspective, an increase in total follistatin in HPD controls relative to non-hpd controls is consistent with the reduced FSH levels (Table 1). Paradoxically, the fall in FSH in HPD controls relative to non-hpd controls was also accompanied by an increase in activin A and B, not a decrease. Earlier studies with sheep have found that activin negatively regulates LH release [55]. The opposing changes in activin and LH following HPD are, therefore, consistent with this premise. Comparison of changes in regulatory proteins in HPD ewes administered GnRH with HPD controls not administered GnRH revealed no effects of altering the frequency of GnRH on circulating levels of activin A and B (Table 2). Both slow- and high-frequency of GnRH pulses increased FSH, but this increase was not accompanied by an increase in stimulatory input (activin) or decrease in total follistatin at the peripheral level. The trend for an increase in activin B in HPD control relative to non-hpd controls (Table 1) and the lack of regulation by GnRH (Table 2) are suggestive of hypothalamic inhibition of circulating activins. Taking all regulators in concert, the endocrine input to pituitary FSH from activins and follistatins appears to be very complex. Pituitary level. The findings from this study, while confirming production of inhibin A, activin B, and follistatin within the pituitary, reveal a complex interaction between hypothalamic GnRH and the paracrine environment (Tables 1 and 2). The main GnRH-induced change that was evident at the pituitary level was with total follistatin and immunoreactive inhibin (not seen with dimeric inhibin A; Table 2). No changes in activin B were evident with any manipulations. An increase in total pituitary follistatin levels with 3-hourly but not hourly pulses of GnRH relative to HPD control (Table 2) is also at odds with earlier findings in rats, that high- but not lowfrequency GnRH increased follistatin mrna [3]. The differences between the present study and the Kirk et al. [3] study may relate to measures in the present study being at the protein level and the latter at the mrna level. In the context of overall pituitary paracrine tone, contrary to what was expected, the increase in pituitary FSH in the 3-hourly GnRH group was associated with an increase, not a decrease, in total follistatins. This is paradoxical, as follistatins are negative regulators of FSH [7, 8]. A likely possibility is that FSH regulation is under autocrine control and that cell-specific changes are masked by measurement in whole pituitary. Absence of activin A in pituitary suggests circulating levels of activin A come from sites other than the pituitary; activin A is produced at various sites within the body [50, 51]. Any paracrine role for activin in influencing FSH synthesis/secretion should be therefore mediated via activin B rather than activin A. This is supported by earlier findings that neutralizing antiserum directed against the activin bb-subunit but not ba-subunit profoundly suppresses FSH secretion from rat pituitary cells in culture [14, 15]. In addition to addressing the impact of GnRH pulse frequency on the FSH regulators, this study demonstrates the powerful combination of HPD and monitoring GnRH in hypophyseal portal circulation to assess delivery patterns of exogenous GnRH at the pituitary level. Comparison of the endogenous pattern of GnRH in portal blood of ovariectomized ewes with that of administered GnRH in HPD ovariectomized ewes revealed that the disappearance of exogenous GnRH in hypophyseal portal blood was slower than the endogenously produced GnRH. The pattern of administered GnRH in this study has been noted in the peripheral circulation of ovariectomized ewes (clamped with estradiol and progesterone to block GnRH endogenous pulses) following pulsatile administration of GnRH [56]. The more rapid disappearance of endogenous GnRH pulses was expected given that endogenous pulses are measured in hypophyseal portal circulation only during their first pass, and they get rapidly diluted in the peripheral circulation, where they become immeasurable. Larger amounts of GnRH need to be administered peripherally to enable sufficient concentrations of GnRH to reach the pituitary. While the disappearance of administered GnRH reflects the circulating half-life of GnRH [57], the disappearance of endogenous GnRH from hypophyseal portal blood reflects dilution into the larger body of peripheral blood. Interestingly, the differences in magnitude and duration of endogenous and exogenous GnRH dynamics did not translate to differences in peripheral LH pulse patterns. LH pulse amplitude and total LH released within a pulse were similar between non-hpd ewes and HPD ewes receiving hourly pulses of GnRH. Although unlikely, one possibility is that the delivered GnRH pulses exceed the maximal effective concentrations of GnRH. It is also possible that the LH response to GnRH in non-hpd ewes is potentiated by other hypothalamic factors that would not have influenced pituitary function in HPD ewes. Examples of peptides involved in potentiating response to GnRH include galanin [58] and pituitary adenylate cyclase activating peptide [59]. These findings emphasize the need to exercise caution when deciphering amplitude modulation of GnRH on the basis of LH pulse amplitude. Relative to pituitary LH content, the lack of a drop in pituitary LH content in HPD ewes not treated with GnRH might reflect the lack of LH release in the absence of GnRH coupled with insufficient elapsed time between HPD surgery and tissue collection for pituitary LH stores to be depleted. 7 Article 171

8 SHARMA ET AL. In summary, findings from this study document that GnRH pulse frequency, in addition to being important in regulating FSH production, is important in determining relative amounts of FSH regulatory proteins, specifically follistatin within the pituitary. The findings also emphasize that circulating activin A and B are under hypothalamic inhibition. However, GnRH frequency modulation of total activins, inhibins, and follistatins were unable to account in a straightforward manner for the observed changes in FSH. Specifically, increased FSH production was not accompanied by an increase in stimulatory (activins) or decreased inhibitory (follistatins/inhibins) input to the pituitary, and decreased FSH was not accompanied by reduced stimulatory or increased inhibitory input. Perhaps the effect of GnRH pulse frequency is mediated via changes in the balance of free versus bound amounts of these FSH regulatory proteins (this study measured total concentrations of FSH regulators) or involves other regulators not monitored in this study. REFERENCES 1. Schally AV, Arimura A, Kastin AJ, Matsuo H, Baba Y, Redding TW, Nair RM, Debeljuk L, White WF. Gonadotropin-releasing hormone: one polypeptide regulates secretion of luteinizing and follicle-stimulating hormones. Science 1971; 173: Wildt L, Hausler A, Marshall G, Hutchison JS, Plant TM, Belchetz PE, Knobil E. Frequency and amplitude of gonadotropin-releasing hormone stimulation and gonadotropin secretion in the rhesus monkey. Endocrinology 1981; 109: Kirk SE, Dalkin AC, Yasin M, Haisenleder DJ, Marshall JC. Gonadotropin-releasing hormone pulse frequency regulates expression of pituitary follistatin messenger ribonucleic acid: a mechanism for differential gonadotrope function. Endocrinology 1994; 135: Besecke LM, Guendner MJ, Schneyer A, Bauer-Dantoin AC, Jameson JL, Weiss J. Gonadotropin-releasing hormone regulates follicle-stimulating hormone-beta gene expression through an activin/follistatin autocrine or paracrine loop. Endocrinology 1996; 137: Dalkin AC, Haisenleder DJ, Gilrain JT, Aylor K, Yasin M, Marshall JC. Gonadotropin-releasing hormone regulation of gonadotropin subunit gene expression in female rats: actions on follicle-stimulating hormone beta messenger ribonucleic acid (mrna) involve differential expression of pituitary activin (beta-b) and follistatin mrnas. Endocrinology 1999; 140: Bilezikian LM, Blout AL, Donaldson CJ, Vale WW. Pituitary actions of ligands of the TGF-beta family: activins and inhibins. Reproduction 2006; 132: Ying SY. Inhibins, activins, and follistatins: gonadal proteins modulating the secretion of follicle-stimulating hormone. Endocr Rev 1988; 9: Ling N, DePaolo LV, Bicsak TA, Shimasaki S. Novel ovarian regulatory peptides: inhibin, activin, and follistatin. Clin Obstet Gynecol 1990; 33: Phillips DJ. The activin/inhibin family. In: Thomson A, Lotze MT (eds.), The Cytokine Handbook, vol. 2, 4th ed. London: Academic Press; 2003: Martens JW, de Winter JP, Timmerman MA, McLuskey A, van Schaik RH, Themmen AP, de Jong FH. Inhibin interferes with activin signaling at the level of the activin receptor complex in Chinese hamster ovary cells. Endocrinology 1997; 138: Lebrun JJ, Vale WW. Activin and inhibin have antagonistic effects on ligand-dependent heteromerization of the type I and type II activin receptors and human erythroid differentiation. Mol Cell Biol 1997; 17: Lewis KA, Gray PC, Blount AL, MacConell LA, Wiater E, Bilezikjian LM, Vale W. Betaglycan binds inhibin and can mediate functional antagonism of activin signalling. Nature 2000; 404: Pangas SA, Woodruff TK. Activin signal transduction pathways. Trends Endocrinol Metabol 2000; 11: Corrigan AZ, Bilezikjian LM, Carroll RS, Bald LN, Schmelzer CH, Fendly BM, Mason AJ, Chin WW, Schwall RH, Vale W. Evidence for an autocrine role of activin B within rat anterior pituitary cultures. Endocrinology 1991; 128: Baratta M, West LA, Turzillo AM, Nett TM. Activin modulates differential effects of estradiol on synthesis and secretion of folliclestimulating hormone in ovine pituitary cells. Biol Reprod 2001; 64: Shimonaka M, Inouye S, Shimasaki S, Ling N. Follistatin binds to both activin and inhibin through the common beta-subunit. Endocrinology 1991; 128: Thompson TB, Lerch TF, Cook RW, Woodruff TK, Jardetzky TS. The structure of the follistatin: activin complex reveals antagonism of both type I and type II receptor binding. Dev Cell 2005; 9: Padmanabhan V, West C. Chapter III: endocrine, autocrine, and paracrine actiions of inhibin/activin/follistatin on follicle-stimulating hormone. In: Muttukrishna S, Ledger W (eds.), Inhibin, Activin and Follistatin in Human Reproductive Physiology. London: Imperial College Press, 2001: Padmanabhan V, Karsch FJ, Lee JS. Hypothalamic, pituitary and gonadal regulation of follicle-stimulating hormone. In: Skinner D, Evans N, Doberska C (eds.), Large Mammals as Neuroendocrine Models. Reprod Suppl 2002; 59: Besecke LM, Guendner M, Sluss PA, Polak AG, Woodruff TK, Jameson JL, Bauer-Dantoin AC, Weiss J. Pituitary follistatin regulates activinmediated production of follicle-stimulating hormone during the rat estrous cycle. Endocrinology 1997; 138: Hamernik DJ, Crowder ME, Nilson JH, Nett TM. Measurement of messenger ribonucleic acid for luteinizing hormone beta-subunit, alphasubunit, growth hormone, and prolactin after hypothalamic pituitary disconnection in ovariectomized ewes. Endocrinology 1986; 119: Clarke IJ, Cummins JT, de Kretser DM. Pituitary gland function after disconnection from hypothalamic influences in the sheep. Neuroendocrinology 1983; 36: Clarke IJ, Cummins JT. The temporal relationship between gonadotrophin releasing hormone (GnRH) and luteinizing hormone (LH) secretion in ovariectomized ewes. Endocrinology 1982; 111: Caraty A, Locatelli A, Moenter SM, Karsch FJ. Sampling of hypophyseal portal blood of conscious sheep for direct monitoring of hypothalamic neurosecretory substances. In: Levine JE (ed.), Pulsatility in Neuroendocrine Systems, Methods in Neuroscience, vol. 20. San Diego, CA: Academic Press; 1994: Goodman RL. Neuroendocrine control of the ovine estrous cycle. In: Knobil E, Neill JD (eds.), The Physiology of Reproduction, vol. 2. New York: Raven Press; 1994: Caraty A, Locatelli A, Schanbacher B. Augmentation, by naloxone, of the frequency and amplitude of LH-RH pulses in hypothalamo-hypophyseal portal blood in the castrated ram. C R Acad Sci III 1987; 305: Niswender GD, Reichert LE, Midgley AR, Nalbandov AV. Radioimmunoassay for bovine and ovine luteinizing hormone. Endocrinology 1969; 84: Midgley AR Jr, McFadden K, Ghazzi M, Karsch FJ, Brown MB, Mauger DT, Padmanabhan V. Nonclassical secretory dynamics of LH revealed by hypothalamo-hypophyseal portal sampling of sheep. Endocrine 1997; 6: Padmanabhan V, McFadden K, Mauger DT, Karsch FJ, Midgley AR Jr. Neuroendocrine control of follicle-stimulating hormone (FSH) secretion: I. Direct evidence for separate episodic and basal components of FSH secretion. Endocrinology 1997; 138: Knight PG, Muttukrishna S, Groome NP. Development and application of a two-site enzyme immunoassay for the determination of total activin-a in serum and follicular fluid. J Endocrinol 1996; 148: Tannetta DS, Feist SA, Bleach ECL, Groome NP, Evans LW, Knight PG. Effects of active immunization of sheep against an amino terminal peptide of the inhibin ac subunit on intrafollicular levels of activin A, inhibin A and follistatin. J Endocrinol 1998; 157: Ludlow H, Phillips DJ, Myers M, McLachlan RI, de Kretser DM, Allan CA, Anderson RA, Groome NP, Marko Hyvonen M, Duncan WC, Muttukrishna S. A new total activin B enzyme-linked immunosorbent assay (ELISA): development and validation for human samples. Clin Endocrinol 2009; 71: Robertson DM, Hayward S, Irby D, Jacobsen J, Clarke L, McLachlan RI, de Kretser DM. Radioimmunoassay of rat serum inhibin: changes after PMSG stimulation and gonadectomy. Mol Cell Endocrinol 1988; 58: Knight PG, Feist SA, Tannetta DS, Bleach ECL, Fowler PA, O Brien M, Groome NP. Measurement of inhibin-a (a-b A dimer) during the oestrus cycle, after manipulation of ovarian activity and during pregnancy in ewes. J Reprod Fertil 1998; 113: Bleach EC, Glencross RG, Feist SA, Groome NP, Knight PG Plasma inhibin A in heifers: relationship with follicle dynamics, gonadotropins, and steroids during the estrous cycle and after treatment with bovine follicular fluid. Biol Reprod 2001; 64: Article 171

9 HYPOTHALAMIC CONTROL OF FSH-REGULATORY PROTEINS 36. McNeilly AS, Souza CJH, Baird DT, ISwanston IA, McVerry J, Crawford J, Cranfield M, Lincoln GA. Production of inhibin A not B in rams: changes in plasma inhibin A during testis growth, and expression of inhibin/activin subunit mrna and protein in adult testis. Reproduction 2002; 123: Phillips DJ, Hedger MP, McFarlane JR, Klein R, Clarke JI, Tilbrook AJ, Nash AD, de Kretser DM. Follistatin concentrations in male sheep increase following sham castration/castration or injection of interleukin- 1b. J Endocrinol 1996; 151: Veldhuis JD, Johnson ML. Cluster analysis: a simple versatile, and robust algorithm for endocrine pulse detection. Am J Physiol 1986; 250: E486 E Marshall JC, Dalkin AC, Haisenleder DJ, Griffin ML, Kelch RP. GnRH pulses the regulators of human reproduction. Trans Am Clin Climatol Assoc 1993; 104: Ishizaka K, Kitahara S, Oshima H, Troen P, Attardi B, Winters SJ. Effect of gonadotropin-releasing hormone pulse frequency on gonadotropin secretion and subunit messenger ribonucleic acids in perifused pituitary cells. Endocrinology 1992; 130: Clarke IJ, Cummins JT, Findlay JK, Burman KJ, Doughton BW. Effects on plasma luteinizing hormone and follicle-stimulating hormone of varying the frequency and amplitude of gonadotropin-releasing hormone pulses in ovariectomized ewes with hypothalamo-pituitary disconnection. Neuroendocrinology 1984; 39: Wu FC, Irby DC, Clarke IJ, Cummins JT, de Kretser DM. Effects of gonadotropin-releasing hormone pulse-frequency modulation on luteinizing hormone, follicle-stimulating hormone and testosterone secretion in hypothalamo/pituitary-disconnected rams. Biol Reprod 1987; 37: Haisenleder DJ, Dalkin AC, Ortolano GA, Marshall JC, Shupnik MA. A pulsatile GnRH-releasing hormone stimulus is required to increase transcription of the gonadotropin subunit genes: evidence for differential regulation of transcription by pulse frequency in vivo. Endocrinology 1991; 128: Shupnik MA, Fallest PC. Pulsatile GnRH regulation of gonadotropin subunit gene transcription. Neurosci Biobehav Rev 1994; 18: Gross KM, Matsumoto AM, Southworth MB, Bremner WJ. Evidence for decreased luteinizing hormone-releasing hormone pulse frequency in men with selective elevations of follicle-stimulating hormone. J Clin Endocrinol Metab 1985; 60: Leung K, Kaynard AH, Negrini BP, Kim KE, Maure RA, Landefeld TD. Differential regulation of gonadotropin subunit messenger ribonucleic acids by gonadotropin releasing hormone pulse frequency in ewes. Mol Endocrinol 1987; 1: Finkelstein JS, Badger TM, O Dea L, Spratt DI, Crowley WF. Effects of decreasing the frequency of gonadotropin-releasing hormone stimulation in gonadotropin secretion gonadotropin-releasing hormone-deficient men and perifused rat pituitary cells. J Clin Invest 1988; 81: Adams LA, Clifton DK, Bremner WJ, Steiner RA. Testosterone modulates the differential release of luteinizing hormone and follicle-stimulating hormone that occurs in response to changing gonadotropin-releasing hormone pulse frequency in the male monkey, Macaca fascicularis. Biol Reprod 1988; 38: Welt CK. Regulation and function of inhibins in the normal menstrual cycle. Semin Reprod Med 2004; 22: Knight PG. Roles of inhibins, activins, and follistatin in the female reproductive system. Front Neuroendocrinol 1996; 17: DePaolo LV, Bicsak TA, Erickson GF, Shimasaki S, Ling N. Follistatin and activin: a potential intrinsic regulatory system within diverse tissues. Proc Soc Exp Biol Med 198: Erratum in: Proc Soc Exp Biol Med 1991; 200: Oride A, Kanasaki H, Purwana IN, Mutiara S, Miyazaki K. Follistatin, induced by thyrotropin-releasing hormone (TRH), plays no role in prolactin expression but affects gonadotropin FSHb expression as a paracrine factor in pituitary somatolactotroph GH3 cells. Regul Pept 2009; 156: Winters SJ, Moore JP Jr. PACAP, an autocrine/paracrine regulator of gonadotrophs. Biol Reprod 2011; 84: Shao LE, Frigon NL Jr, Yu A, Palyash J, Yu J. Contrasting effects of inflammatory cytokines and glucocorticoids on the production of activin A in human marrow stromal cells and their implications. Cytokine 1998; 10: Muttukrishna S, Knight PG. Inverse effects of activin and inhibin on the synthesis and secretion of FSH and LH by ovine pituitary cells in vitro. J Mol Endocrinol 1991; 6: Williams CY, Harris TG, Battaglia DF, Viguié C, Karsch FJ. Endotoxin inhibits pituitary responsiveness to gonadotropin-releasing hormone. Endocrinology 2001; 142: Moenter SM, Brand RM, Midgley AR, Karsch FJ. Dynamics of gonadotropin-releasing hormone release during a pulse. Endocrinology 1992; 130: Mechenthaler I. Galanin and the neuroendocrine axes. Cell Mol Life Sci 2008; 65: Counis R, Laverrière JN, Garrel-Lazayres G, Cohen-Tannoudji J, Larivière S, Bleux C, Magre S. What is the role of PACAP in gonadotrope function? Peptides 2007; 28: Article 171