| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Endocrinology, Institute of Pharmacological Sciences (G.E.R., V.C.), University of Milan, Milan, Italy
Address all correspondence and requests for reprints to: Roberto Maggi, Ph.D., Department of Endocrinology, University of Milan, Via G. Balzaretti 9, 20133 Milan, Italy. E-mail: roberto.maggi{at}unimi.it
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In conclusion, these results indicate that immortalized LHRH-secreting neurons express IP receptors through which PGI2 may exert relevant effects on LHRH release.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The LHRH system is composed of only a few hundred neurons, which are not confined to a discrete nucleus but are distributed as clusters extending from the preoptic to the anterior hypothalamic regions (4). Such an organization makes it difficult to determine, by in vivo or ex vivo experiments, whether the different regulatory factors exert their effects directly on the LHRH-synthesizing neurons or whether they act via the activation and/or inhibition of other neuronal systems impinging on the LHRH-secreting neurons (2).
A promising tool for the investigation of the possible excitatory or
inhibitory inputs reaching the LHRH-secreting neurons is represented by
the utilization of immortalized hypothalamic cell lines, such as GT1
(5) and GN (6) cells. Among these, GT1 cells have been extensively
studied, and it has been found that they retain many characteristics of
hypothalamic LHRH-secreting neurons, such as a neuronal morphology, the
expression of neuronal markers, an elevated synthesis of the LHRH gene,
and the typical pulsatile secretion of LHRH (7, 8, 9). In the last few
years, the receptors for several families of neurotransmitters known to
modify LHRH secretion have been identified on GT1 cells, and the
stimulatory or inhibitory activity of the corresponding ligands on LHRH
secretion has been documented (8). In particular, the presence of
functional opioid receptors of the 
-type has been ascertained in
GT1–1 cells, a subclone of the GT1 cell line (10). It has also been
reported that the activation of these receptors leads to an inhibition
of the release of LHRH when this is stimulated by either forskolin or
PGs (11).
Prostanoids (PGs, tromboxanes, and leukotrienes), which derive from arachidonic acid metabolism, have been shown to affect reproductive functions (12, 13, 14, 15). Among the various PGs, PGE2 appears to be the one mainly involved in the control of LHRH secretion in vivo (16, 17). PGE2 is also able to induce a significant release of LHRH from GT1 cell line (11, 18, 19), an effect accompanied by an increase in intracellular cAMP levels and probably mediated by the interaction with PG receptors of the EP series (see below) (20). However, the involvement of PGE1 in the control of the secretion of LHRH in vivo has also been reported (13, 21), and previous results obtained in the authors’ laboratory have shown that PGE1 is more potent than PGE2 in stimulating both LHRH release and cAMP accumulation in GT1–1 cells (10, 11). According to the most recent classification of the PG receptors (22), PGE1 appears to interact with the putative PGI2 receptor (IP receptor) with an affinity and a specificity higher than those found for any other member of the PG receptor family (e.g. EP).
The strong effect exerted by PGE1 on the release of LHRH from GT1–1 cells supports the hypothesis that IP receptors might be present on these cells, and that their activation might influence the secretion of LHRH. This hypothesis has been verified by evaluating the expression of IP receptors in two immortalized LHRH neuronal cell lines (GT1–1 and GN11) using RT-PCR. Moreover, GT1–1 cells were used to perform specific binding assays for IP receptors and to measure the intracellular accumulation of cAMP and the release of LHRH after treatment with PGI2, its synthetic analogs iloprost and cicaprost, PGE1, and PGE2.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-32P]deoxy-ATP were purchased from Amersham
(Aylesbury, UK). All other analytical grade reagents were obtained from
Merck (Darmstad, Germany) and Sigma Chemical Co.
Cell cultures
GT1–1 and GN11 cells were provided by R. I. Weiner (San
Francisco, CA) and S. Radovick (Children’s Hospital, Boston, MA),
respectively. The cells were routinely grown in monolayer at 37 C in a
humidified CO2 incubator in DMEM containing 1
mM sodium pyruvate, 100 µg/ml streptomycin, 100 U/ml
penicillin, and 10 mg/liter phenol red and supplemented with 10% FBS.
The medium was replaced at 2-day intervals. Subconfluent cells were
routinely harvested by trypsinization and seeded in 57-cm2
dishes (1 x 106 cells). GT1–1 and GN11 cells within
six passages were used throughout the experiments.
RT-PCR analysis
Expression of the specific messenger RNA (mRNA) coding for mouse
IP receptors was tested after nucleic acid amplification, using a PCR
procedure (RT-PCR), of reverse transcribed total RNA extracted from
GT1–1 and GN11 cells. Cultured cells were solubilized in Ultra-Spect
reagent (Biotex Laboratories, Inc., Houston, TX)
and processed for total RNA extraction according to the protocol
included in the reagent kit. RT and PCR were performed using a GeneAmp
kit (Perkin Elmer, Branchburg, NJ) on 2 µg total RNA
from each sample. Samples of total RNA of adult mouse spleen, known to
contain high levels of IP receptor mRNA (23), were obtained as
described for the cells and used as positive controls. All of the
synthetic oligonucleotides, obtained from Pharmacia Biotech (Uppsala, Sweden), were deduced from the complementary
DNA (cDNA) sequence of the mouse IP receptor (GenBank accession no.
D26157) and verified with the program Amplify (Bill Engels, University
of Wisconsin, Madison, WI). The synthetic oligonucleotides used as
primers were designed to amplify sequence 500-1128 of the IP cDNA, and
their sequences were as follows: upstream primer,
5'-CAC.CCA.TCG.GCC.TTT.GCG.GT-3'; and downstream primer,
3'-TA.GTA.CCG.GCA.CAC.GAG.GGA-5'. To exclude the presence of
contamination with genomic DNA, a parallel set of the RNA samples was
assayed for RT-PCR, but reverse transcriptase was omitted from the
incubation mixture.
The RT conditions were 42 C for 45 min, followed by 5 min at 95 C, using a final concentration of 1 mM of each deoxy-NTP, 1 U ribonuclease inhibitor, 2.5 U murine leukemia virus reverse transcriptase, and the downstream primer in a final volume of 20 µl. The same buffer (50 mM KCl, 10 mM Tris-HCl, and 2 mM MgCl2) was used for both transcription and amplification. After this step, 2.5 U Ampli-Taq polymerase and the upstream primer were added in a final volume of 100 µl. Samples were amplified by repeated cycles (35 cycles) at 95 C for 1 min, followed by 1 min at 42 C and 1 min at 72 C.
Amplification products were separated by electrophoresis in 2% agarose
gel and detected by ethidium bromide staining. The amplified cDNAs were
transferred to blotting membrane (Hybond Nfp, Amersham) by capillary
elution in 10 x SSC (standard saline citrate) and fixed at 80 C
for 2 h. A synthetic antisense oligonucleotide
(3'-GC.AAG.GCG.GGT.CGG.ACC.GCC-5' complementary to the sequence
893–912 of IP cDNA) was used as radiolabeled probes in Southern
analysis. The labeling reactions were performed at the free 5'-end of
the oligoprimers using T4 DNA polynucleotide kinase, and
[-32P]deoxy-ATP. The cDNAs fixed on the membrane
were incubated at 45 C for 4 h with a prehybridizing solution;
after then, the 32P 5'-end-labeled oligonucleotide probes
(1 x 106 cpm/ml) were added, and the hybridization
reaction was left at 45 C for 16 h. After high stringency washes,
the membranes were exposed to x-ray films (Hyperfilm, Amersham).
RT-PCR blanks were performed using distilled water instead of mRNA and simultaneously subjected to RT-PCR-Southern blotting with the same reagents and conditions described above; in any of the experiments performed, no specific signal could be obtained, indicating that no contamination by any reagent occurred in these studies.
Receptor binding assay
[3H]Iloprost (12.7 Ci/mmol), purchased as a
solution in organic solvent, was evaporated under nitrogen stream, and
the residue was taken up in 10 mM Tris-HCl, pH 8. PGs stock
solutions were diluted in 10 mM Tris-HCl buffer immediately
before use. The program Design (24) was used to optimize the binding
protocols, by selecting the lowest number of most appropriate
concentrations. Homologous iloprost curves were always performed
according to a mixed type protocol, combining both saturation (the
first three concentration points of the curves, 1, 3, and 10
nM) and displacement (the last four concentration points,
from 30 nM to 10 µM) curves (25). By
effectively combining both saturation and competition protocols in a
single curve, one can reach high concentrations of the ligand without
consuming an excessive amount of labeled ligand (competition part of
the curve), yet have adequate radioactivity in the lower concentration
range (saturation part of the curve). Equilibrium binding studies were
performed as previously described (26) with minor variations. Briefly,
GT1–1 membranes (0.1–0.15 mg/sample) were incubated for 5 min at 30 C
in 10 mM Tris-HCl, pH 8 (final volume, 100 µl), with 1,
3, and 10 nM [3H]iloprost and unlabeled
homologous and heterologous ligands at the indicated concentrations.
Unbound ligand was separated from bound ligand by rapid vacuum
filtration onto glass-fiber filters (GF/C, Whatman,
Clifton, NJ), and the filters were washed twice with 2 ml ice-cold 10
mM Tris-HCl. Radioactivity was then measured by liquid
scintillation counting (Filter Count, Packard Instruments Co.,
Meridien, CT). Nonspecific binding was calculated by Ligand (see below)
as one of the unknown parameters of the model and it ranged between
25–30% of the total binding of 10 nM
[3H]iloprost.
Static culture studies
GT1–1 cells were plated in 24-well plates (0.5 x
106 cells/cm2) and used after 5 days of
culture. All of the samples were assayed for protein content using a
microassay (27) with human serum albumin as a standard. No variations
in total protein per well were detected in any of the experimental
groups (data not shown).
On the day of the experiment, GT1–1 cells were washed with 1 ml DMEM (prewarmed at 37 C) before addition of the compounds to be tested, made from 1000-fold concentrated ethanolic solution.
cAMP accumulation.GT1–1 cells were preincubated for 15 min at 37 C in DMEM containing the phosphodiesterase inhibitor IBMX (0.5 mM). Intracellular cAMP accumulation was then measured over a 15-min incubation period at 37 C in DMEM with the various compounds in the presence of IBMX using a commercial assay kit (Amersham) after ethanolic extraction.
LHRH secretion.GT1–1 cells were incubated for 30 min at 37 C in DMEM with the test compounds. At the end of the incubation period the incubation medium was collected and centrifuged for 5 min at 12,000 rpm, and the supernatant was stored at -70 C until RIA. The content of LHRH in the culture medium was determined by RIA using a commercial antibody (L-8391, Sigma Chemical Co.) and iodinated LHRH (Amersham). Synthetic LHRH (NovaBiochem, Laufefingen, Switzerland) was used as standard. All samples from the same experiment were run in duplicate in the same assay. Inter- and intraassay variations were 9.4% and 6.6%, respectively. The sensitivity of the assay was 3.9 pg/ml.
Data analysis and statistical evaluation
Analysis of equilibrium ligand binding data were performed by
means of the computer program Ligand (28). Statistical analysis of the
concentration-response curves was performed using the program Allfit
(29), which calculates the lower and upper plateaux, the slope, and the
EC50 and allows the comparison of two or more curves.
Selection of the best fitting model and evaluation of the statistical
significance of the parameters were based on the F test for the extra
sum of square principle. A statistical level of significance with
P < 0.05 was accepted. All curves shown were computer
generated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows the
autoradiographic data obtained after Southern analysis performed on the
amplification products of the RT-PCR. Lane 1 corresponds to an internal
control, in which no RNA was added to the PCR incubation mixture. The
presence of a specific transcript of the expected size (628 bp) was
detectable in RNA samples obtained from mouse spleen (lane 2), which
was used as the positive control (23). The results also show that a
specific amplification of a transcript for the IP receptor was detected
in both GT1–1 (lane 3) and GN11 (lane 4) cells; no amplification
fragments were obtained from RNA samples when the incubation with RT
was omitted (lanes 5 and 6), a fact that excludes contamination of the
samples with genomic DNA.
|
Specific binding of [3H]iloprost to
GT1–1 cell membrane preparations
Binding experiments were then performed on GT1–1 cell membrane
preparations using the stable PGI2 analog
[3H]iloprost as the ligand for the IP receptors. The
results of homologous binding curves show that
[3H]iloprost bound to GT1–1 membrane preparations in a
dose-dependent manner (0.1 nM to 10 µM; Fig. 2a). Computer analysis of the results
revealed that [3H]iloprost interacts with a single class
of binding sites with an equilibrium dissociation constant
(Kd) of 4.6 nM and a maximal binding capacity
of 85 fmol/mg protein, which corresponds to about 10,000
sites/cell.
|
. It is evident
that the specific binding of [3H]iloprost on GT1–1 cell
membrane is inhibited by unlabeled PGs in the following order of
potency: iloprost > PGI2 = PGE1 >>
PGE2. In particular, as summarized in Table 1, unlabeled iloprost competes for the
binding of [3H]iloprost with an equilibrium inhibition
constant (Ki) of 4.6 nM;
PGI2 and PGE1 are 14- to 20-fold less potent
than iloprost (Ki = 83.8 and 65.8 nM,
respectively). On the other hand, PGE2 exerts a very weak
competition on [3H]iloprost binding, with a potency
significantly smaller (Ki = 2783.5 nM) than
that of either PGI2 or PGE1. These results
indicate the presence of biologically active IP receptors specific for
PGI2 and its analogs on GT1–1 cell membrane
preparations.
|
The dose-response curves obtained in these experiments are reported in
Fig. 3; because of the differences
observed in maximal responses (see below), the curves are represented
as percent values. It has been found that PGI2 as well as
its synthetic agonists iloprost and cicaprost are similarly potent in
stimulating cAMP accumulation in GT1–1 cells and more potent than
PGE1 and PGE2. The potencies of the different
substances are reported in Table 2; it is
evident that iloprost induces a sustained accumulation of cAMP with an
EC50 of 0.48 nM; PGI2 and its other
stable analog cicaprost are slightly less potent and stimulate cAMP
formation with similar potencies (EC50 1.08 and 1.28
nM, respectively). PGE1 and PGE2
are far less potent than PGI2, with EC50, of
46.8 and 1216.8 nM.
|
|
also shows that the compounds under study exert different
maximal responses on cAMP formation. This parameter corresponds to the
response at the "infinite" dose of each compound, as calculated by
the Allfit program using absolute values. Iloprost appears to be the
most efficient agent, inducing a 49-fold accumulation of cAMP over
basal levels. PGI2 and cicaprost show similar elevated
maximal activities (33- and 38-fold, respectively). PGE1
and PGE2 possess an efficiency significantly lower than
those of PGI2 and its analogs, showing a maximal cAMP
increase corresponding, respectively, to 21 and 8 times the basal
levels.
Effects of different PGs on secretion of LHRH from GT1–1
cells
To verify whether the activation of cAMP-mediated intracellular
pathways elicited by agents stimulating the IP receptor plays a
physiological role in GT1–1 cells, the effects of various PGs on the
release of LHRH were subsequently investigated. The results are
summarized in Fig. 4 and Table 3. The evaluation of LHRH accumulation in
the culture medium under the effects of different PGs shows that
cicaprost exerts the highest potency as an LHRH secretagogue
(EC50 = 0.6 nM), followed by iloprost and
PGI2, which are able to stimulate LHRH release from GT1–1
cells with similar potencies (EC50 = 4.3–4.7
nM). PGE1 is less potent than PGI2
and its stable analogs (EC50 = 28.5); PGE2
shows the lowest potency (215 times less than that of PGI2;
EC50 = 921 nM). The maximal stimulation of LHRH
release evoked by the various PGs is comparable and equivalent to an
increase of 1- to 1.5-fold the basal levels measured in the medium of
unstimulated cells (Table 3).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The experiments devoted to the clarification of the binding characteristics of the IP receptors and their possible physiological roles were performed in GT1 cells, as many mechanisms controlling LHRH release have been well studied in this cell line (8, 31, 32). First, high affinity (Kd = 4.6 nM) binding sites for [3H]iloprost (23, 33) were found in crude GT1–1 cell membrane preparations. The binding characteristics of [3H]iloprost to GT1–1 cells and its specificity are in agreement with those reported for the same tracer when tested on other cells and tissues (33, 34, 35, 36) as well as on Chinese hamster ovary cells transfected with the mouse IP receptor gene (23). Selective competition experiments performed with several PGs further confirm that the binding sites labeled by [3H]iloprost on GT1–1 cells belong to the IP class of PG receptors (22, 23).
It is well known that IP receptors are positively coupled to adenylyl cyclase and that their activation leads to an accumulation of cAMP in several cells and tissues (37). The results presented have clearly shown that PGI2 and its analogs, iloprost and cicaprost, are strong stimulators of cAMP accumulation in GT1–1 cells, whereas PGE2 possesses the lower potency. These data are consistent with previous findings showing that prostacyclin stimulates the intracellular accumulation of cAMP in nonneural and neural cells (23, 30, 38) and indicate, in addition, that this effect is induced via the activation of IP receptors.
The low efficacy of PGE2 on cAMP accumulation was surprising in view of the preliminary identifications in GT1–1 cells of transcripts for the EP1, EP2, and EP3 PG receptors (20) to which PGE2 preferentially binds (22). However, the presence of biologically active receptors of the EP series in GT1–1 cells has yet to be demonstrated by functional or binding experiments.
The functional role of IP receptors detected on GT1–1 cells has been further confirmed in the present study by the finding that prostacyclin and its analogs stimulate the release of LHRH in a dose-dependent manner. In general, this effect reflects the pharmacological profile of the activation of the IP receptors as measured in the other experiments here described (e.g. binding studies and cAMP accumulation). However, even though there is a good correlation between the potencies (EC50) and the efficacies of the various PGs in stimulating cAMP production, this is less evident when cAMP production and LHRH release are analyzed. First, PGI2, iloprost, and cicaprost stimulated cAMP accumulation with similar potencies; however, the latter compound proved to be much more potent than PGI2 and iloprost in stimulating LHRH release. In this context, it is important to remember that the release of LHRH from GT1 cells is under the control of various signaling pathways and that each of these may operate via multiple steps (PKA, PKC, cGMP, ion channels, etc.) (8, 31, 32). It has also been reported that PGE2-stimulated LHRH release involves intracellular calcium (39). Because of these considerations, the apparent divergence in the potencies of cicaprost on cAMP accumulation and LHRH release may be indicative of the activation of signaling systems, other than cAMP (40), induced by cicaprost in GT1–1 cells.
Another point that deserves some comment is the fact that the PGs considered in the present study stimulate cAMP accumulation with different efficacies even if all of the compounds tested show similar maximal LHRH-releasing activity. One possible explanation of this quantitative discrepancy may be that the increase in intracellular levels of cAMP induced by the less effective agent (i.e. PGE2) is already sufficient to totally discharge the pool of releasable LHRH present in GT1–1 cells. On the other hand, the elevated formation of cAMP induced by PGI2 and its analogs might subserve additional cellular functions [for instance, the morphological differentiation of the cells we observed in preliminary experiments (Maggi, R., et al., in preparation)].
If one accepts that immortalized hypothalamic neurons represent a good model for study of the physiological functions of LHRH-producing neurons (8, 9), the obvious conclusion derived from these results is that the selective activation of IP receptors (by PGI2 or by PGE1) may play an important role in the control of LHRH secretion.
The physiological relevance of the presence of IP receptors in and
around the LHRH neuronal system is supported by the identification of
similar amounts of PGE2 and 6-cheto-PGF1
,
the main metabolite of PGI2, in extracts of rat median
eminence (41). In addition, it has been reported that estrogens may
affect the production of PGI2 in different in
vivo and in vitro systems (42).
The data reported here do not permit identification of the possible
source of the PGI2 interacting with the IP receptors
present on hypothalamic LHRH neurons; nevertheless, some hypotheses may
be proposed. The first takes into consideration a possible autocrine
function of PGI2. It has been found that LHRH-liberating
agents (e.g. endothelin) facilitate the release of LHRH from
GT1 cells via stimulation of the production of arachidonic acid
derivatives (18). It is then possible that PGI2 might be
produced by the LHRH neurons themselves in response to stimulatory
signals. A second hypothesis is derived from the recently discovered
involvement of glial cells in the control of LHRH release (43, 44). In
particular, it has been observed that transforming growth factor-
may stimulate the production and release of PGE2 from
astrocytes impinging on the LHRH neurons; in turn, this and other PGs
would induce LHRH release (43). Finally, it may be recalled that at the
level of the median eminence, LHRH axons are in close association with
the endothelium of the pituitary portal vessels, and that some
vasoactive factors released by the blood vessels (e.g.
endothelin and nitric oxide) have been found to affect the release of
LHRH (18, 45). It is then attractive to hypothesize that
PGI2, which is also produced by the vascular endothelium,
might create an additional functional link between the portal vessels
and the hypophysiotropic neurons, leading to a synergizing effect.
In conclusion, the present results indicate that not only PGE2, but also PGs acting on the IP receptors, may participate in the activation of LHRH release. As it is known that PGs represent the final mediator in many biological systems, the present results prompt additional studies on the reciprocal interactions between the most classical neurotransmitter systems known to induce the release of LHRH (norepinephrine, epinephrine, dopamine, nitric oxide, histamine, excitatory amino acids, etc.) and the ubiquitous system of arachidonic acid derivatives.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received April 15, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
receptors in a luteinizing
hormone-releasing hormone producing neuronal cell line. Endocrinology 136:289–295[Abstract]
-opioid agonists in GT1–1 neuronal cells. Endocrinology 136:5177–5181[Abstract]
implants on
luteinizing-hormone release. Endocrinology 100:1585–1594[Abstract]
with the secretion of neuroactive substances that stimulate
the release of luteinizing hormone-releasing hormone. Endocrinology 138:19–25This article has been cited by other articles:
 |
|
 |
|
  W. Yang, V. R. Tuniki, S. Anjaiah, J. R. Falck, C. J. Hillard, and W. B. Campbell Characterization of Epoxyeicosatrienoic Acid Binding Site in U937 Membranes Using a Novel Radiolabeled Agonist, 20-125I-14,15-Epoxyeicosa-8(Z)-Enoic Acid J. Pharmacol. Exp. Ther., March 1, 2008; 324(3): 1019 - 1027. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Maggi, F. Pimpinelli, L. Molteni, M. Milani, L. Martini, and F. Piva Immortalized Luteinizing Hormone-Releasing Hormone Neurons Show a Different Migratory Activity in Vitro Endocrinology, June 1, 2000; 141(6): 2105 - 2112. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. G. Harris, D. F. Battaglia, M. E. Brown, M. B. Brown, N. E. Carlson, C. Viguie, C. Y. Williams, and F. J. Karsch Prostaglandins Mediate the Endotoxin-Induced Suppression of Pulsatile Gonadotropin-Releasing Hormone and Luteinizing Hormone Secretion in the Ewe Endocrinology, March 1, 2000; 141(3): 1050 - 1058. [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 |
), PGE1 (
), and
PGE2 (
). The binding isotherms were analyzed by the
Ligand program, and the binding parameters Bmax,
Kd, and Ki were determined. The data are
representative of two independent experiments performed in duplicate.
), PGE1
(