| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Reproductive Sciences Program and Department of Physiology, University of Michigan, Ann Arbor, Michigan 48109
Address all correspondence and requests for reprints to: Dr. F. J. Karsch, Reproductive Science Program, University of Michigan, 300 North Ingalls Building, Room 1101 SW, Ann Arbor, Michigan 48109-0404. E-mail: fjkarsch{at}umich.edu
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
An immune/inflammatory challenge could suppress reproductive
neuroendocrine activity either by inhibiting GnRH release from the
hypothalamus and/or by inhibiting pituitary responsiveness to the GnRH
stimulus. Most prior work has focused on the hypothalamus as the site
of reproductive neuroendocrine inhibition. Observations in a wide
variety of species indicate that endotoxin or various intermediary
cytokines such as interleukin-1 (IL-1) or tumor necrosis factor-
(TNF
), can inhibit both the pulsatile and surge modes of GnRH
secretion (2, 3, 6, 10, 11, 12, 13, 14, 15). For example, systemic
endotoxin suppresses pulsatile GnRH secretion into pituitary portal
blood of ovariectomized ewes (2), and central delivery of
IL-1 inhibits GnRH release and Fos expression in GnRH neurons at the
time of the preovulatory LH surge of rats (15).
Surprisingly little research has focused on an effect of immune/inflammatory challenge at the level of the pituitary despite knowledge that various inhibitors of reproductive neuroendocrine function can act at this level (gonadal steroids, stress hormones, etc.). An initial study in the rat suggests that central administration of IL-1 may not impair pituitary responsiveness to GnRH (16). More recent work, however, has shown that systemic endotoxin does inhibit the expression of GnRH receptor messenger RNA in the anterior pituitary gland of the rat (17). The latter observation encourages further work to assess the impact of immune/inflammatory challenge on pituitary responsiveness to GnRH.
During the course of examining the suppressive effects of endotoxin on GnRH secretion into pituitary portal blood of the ewe, we observed an unexpected uncoupling between GnRH and LH pulses after endotoxin treatment. In extreme cases, LH pulses were eliminated, whereas GnRH pulses showed either slight or no inhibition. This provided circumstantial evidence that endotoxin may inhibit pituitary responsiveness to GnRH. In this report we first describe this uncoupling of GnRH and LH pulse suppression. Next, we present follow-up experiments performed to test the hypothesis that endotoxin inhibits pituitary responsiveness to GnRH.
| Materials and Methods |
|---|
|
|
|---|
Exp 1: effect of endotoxin on endogenous GnRH and LH
pulsatility
This experiment was conducted during the mid to late breeding
season (November 1997 to January 1998) on 16 ewes ovariectomized 57
months previously. Ewes were surgically fitted with an apparatus for
hypophyseal portal blood collection 24 weeks before sampling. The
surgical preparation and procedure for remote automated sampling of
hypophyseal portal blood have been described by Caraty
(18). This procedure enables sampling of portal blood from
conscious, physiologically noncompromised ewes. Portal blood for assay
of GnRH was drawn continuously from 4 h before to 10 h after
endotoxin and dispensed as 10-min fractions into 0.5 ml ice-cold
bacitracin (3 x 10-3
M). Samples of jugular blood were obtained continuously and
separated into 5- or 10-min fractions for assay of LH and cortisol.
Endotoxin was injected at a dose of 40 ng/kg. This dose is 10-fold less
than that which we use routinely to demonstrate profound suppression of
GnRH and LH pulsatility (2, 19); it is close to the
threshold for suppressing pulsatile LH secretion in ovariectomized ewes
(Armstrong, S., and F. J. Karsch, unpublished observations). For
injection, endotoxin was dissolved as 10 µg/ml in sterile saline on
the day of the experiment, vortexed vigorously, and diluted 1:10 to
provide the 40 ng/kg dose. After sample collection, ewes were killed
with a barbiturate overdose (Beuthanasia, Schering Plough Animal Health Corp., Kenilworth, NJ), and the pituitary was inspected
to confirm appropriate placement of the cut for sampling portal
blood.
Exp 2: effect of endotoxin on pituitary responsiveness to exogenous
GnRH
In this experiment we used a model in which endogenous GnRH
pulses were blocked in ovariectomized ewes. Blockade of GnRH pulses was
achieved by delivery of ovarian steroids to produce a midluteal phase
serum level of either progesterone (Run 1) or estradiol (Run 2) during
the anestrous season. These treatments were previously found to
eliminate endogenous GnRH pulsatility (20, 21, 22). For the
purpose of this study, this steroid-block model was preferable to one
involving surgical ablation of GnRH secretion (e.g.
hypothalamo-pituitary disconnection or hypothalamic lesion). The latter
approach would compromise other hypothalamic functions, such as stress
axis activation, which might directly or indirectly influence
gonadotropin secretion after endotoxin treatment. The experiment was
conducted in two runs during the anestrous season (MayJuly of 1998
and 1999). Both runs were conducted as a cross-over design in which
each ewe acted as her own control. The design of both runs is
illustrated in Fig. 1
.
|
Run 2 (Fig. 1
, bottom). Run 1 suggested that
endotoxin inhibits pituitary responsiveness to GnRH. The results,
however, were not definitive because the progesterone treatment failed
to produce the expected complete blockade of endogenous GnRH/LH
pulsatility in some ewes. Further, in some ewes receiving saline
instead of endotoxin, responses to the hourly GnRH injections tended to
increase gradually during the 15-h observation period. This suggested
the 12-h priming period might have been insufficient to stabilize
pituitary responsiveness to the exogenous GnRH pulses. The second run
of the experiment was thus modified in two ways: the priming period was
extended to 4 days, and estradiol, rather than progesterone, was used
to block endogenous GnRH pulses. Estradiol is an extremely potent
inhibitor of pulsatile GnRH secretion in the anestrous ewe
(22). In addition, the goal of Run 2 was expanded to
determine whether endotoxin inhibits pituitary responsiveness to GnRH
via a PG-dependent mechanism. This was prompted by recent evidence that
PGs mediate the suppressive effects of endotoxin on pulsatile GnRH and
LH secretion in the ewe (19).
Run 2 was conducted on eight ewes ovariectomized 12 months previously.
Before the study, each ewe received a 3-cm estradiol implant that
maintains a luteal phase concentration of circulating estradiol (
2
pg/ml) (25). A cannula was inserted into each jugular
vein, one for GnRH delivery and one for blood collection. GnRH (5
ng/kg, 3-ml injection volume, fresh solution prepared daily) was
administered hourly as 6-min pulses using an infusion pump (MiniPuls 2,
Gilson, Middletown, WI) controlled by an electronic timer (model
6227, Fisher Scientific, Pittsburgh, PA). After a 4-day
priming period, endotoxin or saline vehicle was delivered as an iv
bolus (400 ng/kg, 0 h in Fig. 1
). In addition, each ewe received
flurbiprofen (2 mg/kg, iv;
1.0-ml injection volume;
Sigma) or its vehicle (95% ethanol) 30 min before
endotoxin (or saline) and again 5 h later. Flurbiprofen inhibits
the cyclooxygenase-1 and -2 enzymes required for PG synthesis
(27). The dose selected was previously found to reverse
the inhibition of pulsatile GnRH and LH secretion in ovariectomized
ewes and to block endotoxin-induced fever (19), which is a
PG-dependent response (28, 29).
Run 2 was conducted as a cross-over experiment in which each ewe
received three treatments: 1) vehicle for both endotoxin and
flurbiprofen (control), 2) endotoxin plus vehicle for flurbiprofen, 3)
endotoxin plus flurbiprofen. [Note, flurbiprofen alone was not given
as this treatment does not affect LH pulses in ovariectomized ewes
(19).] Blood was sampled at 10-min intervals from 6
h before endotoxin or vehicle until 12.5 h after endotoxin to
monitor pituitary responsiveness to GnRH. To quantify the
experimentally produced GnRH concentrations, aliquots of the jugular
samples surrounding six GnRH pulses (24 h before and 46 h after
endotoxin/vehicle, shaded boxes in Fig. 1
) during one run of
the cross-over experiment were dispensed into bacitracin and frozen for
subsequent GnRH assay. Rectal temperature was taken hourly to assess
efficacy of the endotoxin challenge and to confirm that flurbiprofen
blocked fever.
Assays
LH was measured in duplicate aliquots of plasma (10200 µl)
using a modification (30) of a previously described RIA
(31, 32). Mean intra- and interassay coefficients of
variation (33 assays) were 6.5% and 6.9%, respectively, and assay
sensitivity for 200 µl averaged 0.8 ng/ml NIH LH-12. For GnRH assay,
samples were centrifuged within 1 h of collection to remove cells,
and plasma was stored at -80 C until assay. A 750-µl aliquot of
sample (
600 µl plasma and 150 µl bacitracin) was extracted in 2
ml methanol. Duplicate aliquots of extract (
240 µl portal plasma)
were assayed for GnRH using a previously described RIA (33, 34). In Exp 1, all GnRH values for a given ewe were determined
in a single assay; in Exp 2, all GnRH values were determined in one
assay. Intraassay variation, assessed as median variance ratio of assay
replicates (35), averaged 6.8%, and assay sensitivity was
0.2 pg/tube. Plasma cortisol concentrations were determined in
duplicate 50-µl aliquots of unextracted plasma using the Coat-A-Count
cortisol assay kit (Diagnostic Products, Los Angeles, CA),
validated for use in sheep (2). Modifications of the kit
protocol included adding a low point on the standard curve and
increasing the incubation time with radioactive tracer to 1 h at
37 C. Mean intra- and interassay coefficients of variation (nine
assays) were 5.9% and 6.4%, respectively. Sensitivity averaged 0.8
ng/ml.
Data analysis
In Exp 1, GnRH in pituitary portal blood was assessed as a
collection rate (picograms per min) rather than a concentration. This
minimizes error due to contamination of portal samples with peripheral
blood or cerebrospinal fluid or due to changes in pituitary blood flow.
Formal pulse analysis and statistical comparisons were not performed
because this was an observational study conducted for another purpose,
and controls treated with vehicle instead of endotoxin were not
included.
In Exp 2, all exogenous GnRH pulses induced an increase in circulating LH. In Run 1, amplitudes of these LH responses (peak minus preceding nadir) were averaged across three 5-h periods: one preendotoxin/vehicle (-5 to 0 h) and two postendotoxin/vehicle (05 and 510 h). The percent change in mean amplitude of the LH response between pre- and posttreatment periods was calculated in each ewe for both the endotoxin and vehicle runs of the cross-over study. [The 05 h period was removed from the analysis due to a technical problem with the animal model (see Results).] To determine if endotoxin or vehicle altered pituitary responsiveness to GnRH pulses, the percent change in LH pulse amplitude within each ewe (pretreatment value vs. value 510 h after endotoxin/vehicle) was compared by paired t test. In Run 2, LH pulse amplitudes during the 18-h experiment were averaged across six 3-h periods: two periods during preendotoxin/vehicle treatment and four periods posttreatment. Next, the percent change across the observational period was calculated by determining the ratio of the mean LH pulse amplitude in the first pretreatment period to that in each subsequent period. These ratios were log transformed and analyzed by repeated measures ANOVA in which both treatment and time were repeated measures. To identify where significant interactions existed among treatments, each treatment was deleted in turn, and the analysis was repeated in a pairwise manner. This determined which treatments differed across time.
| Results |
|---|
|
|
|---|
|
0.05;
Fig. 3C
|
|
|
|
0.001, by
paired t test on mean peak values -5 to -2 h
vs. 4 to 6 h; Fig. 7
|
|
| Discussion |
|---|
|
|
|---|
In Exp 2 we directly tested the hypothesis that endotoxin inhibits pituitary responsiveness to exogenous GnRH pulses, using ovariectomized sheep in which endogenous GnRH secretion was blocked by ovarian steroids. Note, in this experiment we used a higher dose of endotoxin (400 ng/kg), as it is the dose we have typically used to inhibit both GnRH and LH pulsatility in ovariectomized ewes (2, 19). Additionally, prior results describing the time course of GnRH and LH suppression suggested that this dose may inhibit pituitary responsiveness to GnRH (2).] The results of Run 1 supported the hypothesis that endotoxin reduced the amplitude of the LH response to exogenous GnRH pulses, although technical difficulties precluded definitive conclusion. These difficulties were overcome in Run 2, and the results demonstrated conclusively that endotoxin inhibited pituitary responsiveness to GnRH.
It is important to consider our findings in the context of the prior conclusion that the pituitary response to GnRH in the rat is not suppressed by central delivery of IL-1, one component of the cytokine cascade that mediates responses to immune/inflammatory challenge (16). Although seemingly at odds with the present conclusion, the design of the two studies differed substantially: vastly different GnRH doses, single vs. multiple GnRH injections, assessment of response to GnRH at one vs. multiple time points, etc. Most notably, the immune/inflammatory challenges themselves differed: a specific cytokine delivered centrally in the prior study and peripheral endotoxin used here. In essence, the prior study in rats is not comparable to that reported here, and it does not weaken our present conclusion that endotoxin inhibits pituitary responsiveness to GnRH.
Our study indicates the effect of endotoxin on responsiveness to GnRH
is evident regardless of whether the pituitary is stimulated by
endogenous or exogenous GnRH pulses (Exp 1 and 2, respectively).
Nevertheless, from the data presented it might seem as if endotoxin had
a more potent inhibitory effect when endogenous GnRH pulses stimulated
the pituitary (compare Figs. 2
and 5
). It is important to point out,
however, that the two experiments were not strictly comparable, because
different doses of endotoxin were employed, and the exogenous GnRH
pulses in Exp 2 did not fully replicate those produced endogenously in
Exp 1. In addition, the animal models differed in the two experiments,
ovariectomized ewes in Exp 1 and steroid-treated ovariectomized ewes in
Exp 2. As discussed below, ovarian steroids modify responses to immune
challenge. Thus, a quantitative comparison between responses in the two
experiments is not appropriate. The main point is that endotoxin
compromises pituitary responsiveness regardless of whether endogenous
or exogenous GnRH pulses stimulate the pituitary.
The present demonstration that endotoxin inhibits pituitary responsiveness to GnRH raises questions related to the mechanisms that mediate this effect. Recent work in rodents has shown that systemic endotoxin decreases GnRH receptor gene expression in the pituitary (17). This effect was seen within 3 h, a time frame consistent with the endotoxin-induced suppression of pituitary responsiveness that we observed in the ewe. We can hypothesize, therefore, that endotoxin reduces GnRH receptor expression in sheep, thereby suppressing the responsiveness of the gonadotrope to GnRH. Further work is required to test this hypothesis and to assess other mechanisms within the pituitary that may mediate the response to this immune challenge.
It is also important to assess whether endotoxin acts directly upon the
pituitary to exert these effects and whether it acts via the same
intermediates that generate other pathophysiological responses to
immune/inflammatory stimuli. At the level of GnRH secretion, endotoxin
stimulates a cascade of peripheral and central cytokines such as IL-1
and TNF
that, together with enhanced synthesis of PGs and perhaps
hormones of the neuroendocrine stress axis, inhibit GnRH release
(7, 8, 9, 10, 11, 19, 37, 38). The question arises of whether the
same mediators suppress LH secretion at the level of the pituitary.
Recent evidence indicates that endotoxin enhances pituitary expression
of IL-1, TNF
, and both their receptor messenger RNAs
(39), but whether these cytokines act locally within the
pituitary to suppress responsiveness to GnRH remains to be tested. With
regard to PGs, we recently demonstrated that administration of a PG
synthesis inhibitor, flurbiprofen, abolished endotoxin-induced
suppression of pulsatile GnRH and LH secretion in the ovariectomized
ewe (19). Here, we tested whether flurbiprofen would also
reverse endotoxin-induced suppression of pituitary responsiveness to
GnRH. To our surprise, flurbiprofen was ineffective in this regard,
although it did abolish the generation of fever, suggesting that PG
synthesis was effectively blocked.
One interpretation of this finding is that PGs do not mediate
endotoxin-induced suppression of pituitary responsiveness. There is,
however, another explanation, one related to the presence or absence of
estradiol. Our prior observation that flurbiprofen prevented
endotoxin-induced LH suppression was obtained in ovariectomized ewes
not replaced with ovarian steroids (19). Of considerable
interest, we recently obtained preliminary evidence that flurbiprofen
failed to reverse endotoxin-induced LH suppression in ovariectomized
ewes treated with estradiol (Breen, K. M., and F. J. Karsch,
unpublished data). Importantly, in the present study the response to
flurbiprofen was tested in ovariectomized ewes treated with estradiol.
Substantial evidence demonstrates that estradiol influences responses
to immune/inflammatory stimuli. For example, IL-1 inhibits LH secretion
in ovariectomized monkeys, but estradiol protects against this
inhibitory effect (11, 40). Estradiol regulates production
of the proinflammatory cytokines IL-1, IL-6, and TNF
(41) and influences the extent to which endotoxin
stimulates cytokine synthesis (42). Collectively, these
findings not only suggest that estradiol regulates the response to
immune challenge, but also that it does so by modulating underlying
intermediary pathways and possibly dependence upon PGs. Thus, we
suggest that in the presence of estradiol, a pathway not involving PG
synthesis mediates endotoxin-induced suppression of pituitary
responsiveness to GnRH.
Although this putative intermediary pathway remains to be identified, we are intrigued by the possibility that it may involve cortisol. Recent work indicates that an elevation of cortisol to levels that we observed during endotoxin challenge inhibits expression of the GnRH receptor and suppresses pulsatile LH secretion in gonadectomized sheep, provided that they are treated with estradiol (43, 44). In the absence of estradiol, cortisol is ineffective. Although the acute stimulation of cortisol does not appear to be necessary for endotoxin-induced inhibition of LH pulsatility in ovariectomized ewes (45), a mediatory role for cortisol in the presence of estradiol has not yet been tested. Further experiments are thus warranted to determine whether, in the presence of estradiol, endotoxin inhibits pituitary responsiveness to GnRH via a PG-independent mechanism, possibly one involving cortisol.
In summary, this study provides exciting new evidence that the pituitary is an important site for reproductive inhibition in response to immune/inflammatory challenge. Specifically, we observed that systemic endotoxin suppresses pituitary responsiveness to GnRH. It is now important to ascertain whether this inhibition at the level of the pituitary is influenced by the ovarian steroid milieu, if this inhibitory effect changes during the course of the ovulatory cycle, and what mediators are involved. Such studies would help clarify mechanisms for the well documented disruptive effects of immune/inflammatory challenge on ovarian cyclicity.
| Acknowledgments |
|---|
| Footnotes |
|---|
2 Present address: INSERM, U-501, IFR Jean Roche,
Faculté Nord, boulevard Pierre Dramard, 13016 Marseilles Cedex
16, France. ![]()
Received September 19, 2000.
| References |
|---|
|
|
|---|
mediates endotoxin induced suppression of
gonadotropin-releasing hormone pulse generator activity in the rat.
Endocr J 44:141148[Medline]
on basal
gonadotropin release in the ovariectomized rhesus monkey: reversal by a
corticotropin-releasing factor antagonist. Endocrinology 128:20772082
acts at the level of the brain and the gonads to interfere with
gonadotropin and sex steroid secretion. Endocrinology 124:21052109
. J Reprod Dev 43:181187[CrossRef]
are produced in the mouse
uterus during the estrous cycle and are induced by estrogen and
progesterone. Dev Biol 151:297305[CrossRef][Medline]
This article has been cited by other articles:
![]() |
R. F. Cooke, J. D. Arthington, D. B. Araujo, and G. C. Lamb Effects of acclimation to human interaction on performance, temperament, physiological responses, and pregnancy rates of Brahman-crossbred cows J Anim Sci, December 1, 2009; 87(12): 4125 - 4132. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. R. Wagenmaker, K. M. Breen, A. E. Oakley, B. N. Pierce, A. J. Tilbrook, A. I. Turner, and F. J. Karsch Cortisol Interferes with the Estradiol-Induced Surge of Luteinizing Hormone in the Ewe Biol Reprod, March 1, 2009; 80(3): 458 - 463. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. R. Wagenmaker, K. M. Breen, A. E. Oakley, A. J. Tilbrook, and F. J. Karsch Psychosocial Stress Inhibits Amplitude of Gonadotropin-Releasing Hormone Pulses Independent of Cortisol Action on the Type II Glucocorticoid Receptor Endocrinology, February 1, 2009; 150(2): 762 - 769. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. M. Breen, A. E. Oakley, A. V. Pytiak, A. J. Tilbrook, E. R. Wagenmaker, and F. J. Karsch Does Cortisol Acting Via the Type II Glucocorticoid Receptor Mediate Suppression of Pulsatile Luteinizing Hormone Secretion in Response to Psychosocial Stress? Endocrinology, April 1, 2007; 148(4): 1882 - 1890. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Xiao, L. Xia-Zhang, N. Vulliemoz, J. Rivier, and M. Ferin Astressin B, a Corticotropin-Releasing Hormone Receptor Antagonist, Accelerates the Return to Normal Luteal Function after an Inflammatory-Like Stress Challenge in the Rhesus Monkey Endocrinology, February 1, 2007; 148(2): 841 - 848. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. A. Daniel, T. H. Elsasser, A. Martinez, B. Steele, B. K. Whitlock, and J. L. Sartin Interleukin-1{beta} and tumor necrosis factor-{alpha} mediation of endotoxin action on growth hormone Am J Physiol Endocrinol Metab, October 1, 2005; 289(4): E650 - E657. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. M. Breen, C. A. Stackpole, I. J. Clarke, A. V. Pytiak, A. J. Tilbrook, E. R. Wagenmaker, E. A. Young, and F. J. Karsch Does the Type II Glucocorticoid Receptor Mediate Cortisol-Induced Suppression in Pituitary Responsiveness to Gonadotropin-Releasing Hormone? Endocrinology, June 1, 2004; 145(6): 2739 - 2746. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. M. Breen and F. J. Karsch Does Cortisol Inhibit Pulsatile Luteinizing Hormone Secretion at the Hypothalamic or Pituitary Level? Endocrinology, February 1, 2004; 145(2): 692 - 698. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. M. Breen, H. J. Billings, N. Debus, and F. J. Karsch Endotoxin Inhibits the Surge Secretion of Gonadotropin-Releasing Hormone via a Prostaglandin-Independent Pathway Endocrinology, January 1, 2004; 145(1): 221 - 227. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Debus, K. M. Breen, G. K. Barrell, H. J. Billings, M. Brown, E. A. Young, and F. J. Karsch Does Cortisol Mediate Endotoxin-Induced Inhibition of Pulsatile Luteinizing Hormone and Gonadotropin-Releasing Hormone Secretion? Endocrinology, October 1, 2002; 143(10): 3748 - 3758. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |