| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Laboratory of Cell Pharmacology, University of Leuven, Medical School, Campus Gasthuisberg, B3000 Leuven, Belgium
Address all correspondence and requests for reprints to: Professor Carl Denef, Laboratory of Cell Pharmacology, University of Leuven, Medical School, Campus Gasthuisberg (O and N), B-3000 Leuven, Belgium. E-mail: carl.denef{at}med.kuleuven.be.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-methyl 5-HT; 5-carboxytryptamine (agonist at 5-HTR1, -5, and -7); tryptamine (preferential 5-HTR7 agonist); and the selective 5-HTR1B agonist CP93129 but not the 5-HTR1A agonists 7-(dipropylamino)tetralin-1-ol-8-hydroxy-2-(di-n-propylamino)tetralin and the 5-HTR1B/D agonist sumatriptan. The selective 5-HTR2B agonist BW 723C86 stimulated GH release, and the selective 5-HTR2B antagonist SB-204741 attenuated the GH response to 5-HT. The present data suggest that 5-HT may release GH through a pituitary site of action, and that the 5-HTR2B, 5-HTR7 and 5-HTR1B mediate this response, with possibly an inhibitory component of the 5-HTR1D. The relative contribution of these receptors may be modulated by estrogen. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
5-HT affects GH release through the serotoninergic projections from the medial and dorsal raphe nucleus of the brain stem to the hypothalamus (4). In vivo-manipulated 5-HT neurotransmission or treatment with serotonin agonists or antagonists suggests a stimulatory, inhibitory, no role, or controversial findings of 5-HT in GH release, depending on the physiological condition and species (5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19). The disparity in findings is probably related to the complexity of the serotoninergic system and the many receptor subtypes involved. Little is known concerning the receptors involved in 5-HT-modulated GH release because the above studies were conducted more than 2 decades ago, when selective 5-HTR agonists or antagonists were not available yet. At most, some data have shown the involvement of the 5-HTR1B/D, 5-HTR2A/C, and 5-HTR3 subtypes (20, 21, 22, 23, 24, 25).
There is some evidence that 5-HT may also affect GH secretion at the level of the pituitary. One group of investigators showed that 5-HT is capable to stimulate GH release in intact rat pituitary in vitro (26). An in vitro effect was seen by another group but only when the posterior pituitary was coincubated (27). However, to our knowledge, the 5-HTR involved has not been studied. The 5-HTR6 is expressed at the mRNA level in the normal pituitary, but its function remains unknown (28). Although there may be an intrinsic hypothalamic serotoninergic system (29) and 5-HT has been detected in the portal blood, the concentration in the latter was not higher than in peripheral blood (30), suggesting no delivery of 5-HT from the median eminence to the anterior pituitary. Basal plasma levels of 5-HT in rats and humans have been reported to be 0.6–0.9 nM but can rise 50–100 times during treatment with amphetamines and selective serotonin reuptake inhibitors (31, 32) or under certain pathological conditions such as hypoxia (32, 33). There are some indications that 5-HT may be contained within the pituitary. Three decades ago, it was shown that the rat pituitary ex vivo is able to convert tryptophan into 5-HT (34), although it remains unknown whether the enzyme is located in glandular cells. Others demonstrated 5-HT in nerve fibers of the neurohypophysis (35) and the intermediate (36, 37) and anterior lobe (38). Interestingly, synapse-like structures between endocrine cells and nerve fibers have been demonstrated in the intermediate (36) and anterior lobes (39, 40). 5-HT has been identified in the secretory granules of gonadotrophs in vivo (35, 41, 42), and it is taken up by anterior pituitary cells via a fluoxetine-sensitive transport system, suggesting expression of the serotonin transporter (SERT) (41, 43). Because 5-HT may be present in the pituitary, the monoamine may influence hormone secretion by a local paracrine or neurocrine action.
The paucity of clear-cut data concerning the GH secretory response to 5-HT at the pituitary vs. hypothalamic level and the current availability of several selective 5-HTR antagonists and certain selective agonists prompted us to investigate whether 5-HT is capable of affecting basal GH secretion in reaggregated anterior pituitary cell cultures (1, 44) and to characterize the receptors behind any intrinsic activity. We report that 5-HT acutely stimulates GH release and present pharmacological evidence for the implication of the 5-HTR2B, 5-HTR7, 5-HTR1B, and possibly the 5-HTR1D. The magnitude of the response was increased by E2, linking 5-HT to putative sexual dimorphic functions of the pituitary.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-methyl 5-HT, m-chlorophenyl piperazine (mCpp), RS-102221, and 7-(dipropylamino)tetralin-1-ol-8-hydroxy-2-(di-n-propylamino)tetralin (8-OHDPAT), BW 723C86, and CP193129 were purchased from Tocris Biosciences (Avonmouth, UK). T3, dexamethasone (Dex), and BSA (Cohn fraction V) was from Serva (Heidelberg, Germany). E2 and vitamin C were from Merck (Darmstadt, Germany). 5-HT was dissolved in 1% vitamin C solution and made fresh. Domperidone, E2, Dex, and ketanserin were dissolved in ethanol and stored at 4 C. IBMX was made fresh and dissolved in dimethylsulfoxide. GR113808 was dissolved in dimethylsulfoxide at a stock concentration of 10–2 M. All other chemicals (analytical grade) were dissolved in water and stored at –20 C. Stock concentrations were 10–3 M, except for Dex and E2 (stock concentration 10–5 M) and T3 (10–7 M stock concentration).
Animals
Adult male Wistar rats (3 months old) were obtained from Elevage Janvier (Schaijk, The Netherlands) and housed in an environment with constant temperature, humidity, and day-night cycles, with food and drink ad libitum. All experiments were conducted in accordance with the Guidelines for Care and Use of Experimental Animals and approved by the university ethical committee.
Anterior pituitary aggregate cell culture
Rats were decapitated after CO2 anesthesia, and the anterior pituitary was dissected free from the neurointermediate lobe under sterile conditions and dispersed into single cells as described previously (1, 44). Dispersed cells were allowed to aggregate at 2 x 106 cells in nontreated 35-mm petri dishes (Iwaki, Japan) with 2 ml serum-free defined culture medium. For aggregation the dishes were placed on a gyratory shaker (AppliTek, Nazareth, Belgium) at 64 rpm in a 1.5% CO2 humidified incubator at 37 C. Culture medium was serum-free defined medium as previously described (44, 45). It consists of HEPES/TES-buffered DMEM/F12 1:1 with selenite and ethanolamine (prepared as powder by Invitrogen, Grand Island, NY), supplemented with 0.5% BSA (endotoxin-poor; Serva), 5 mg/liter insulin-Zn, 5 mg/liter transferrin, 50 mg/liter streptomycin, 35 mg/liter penicillin, 10 mM ethanol, 1 mg/liter catalase, 8 mg/liter phenol red, and 1 g/liter NaHCO3. Final concentration of iron (Fe2+ and Fe3+) was 0.6 µM and glucose 1.4 g/liter (7.8 mM). Medium was replaced after 2 d and aggregates were tested on d 5 of culture. Depending on the experimental design, the medium was either supplemented or not with 1 nM E2, 1 nM T3, or 10 nM Dex during the entire culture period. In certain experiments, pertussis toxin (100 ng/ml; Sigma) was added in the 35-mm petri dishes containing the aggregates for a period of 24 h before the perifusion.
Perifusion of aggregates
The perifusion system was designed to study the continuous release of hormones from aggregates over a period of time, with the possibility of adding substances and observing their effect on hormone secretion at intervals of 1–2 min. A more elaborate description of its design and set-up can be found elsewhere (44). We have thoroughly validated the system and used it in numerous studies on secretory responses to GnRH and TRH, adrenergic agonists, GHRH and vasoactive intestinal peptide (VIP), cholinomimetics, dopamine, angiotensin II (44), bombesin (46), and prolactin (PRL)-releasing peptide (47). Briefly, aggregates (2 x 106 cells) are placed on a nylon mesh, in a chamber that is kept at 37 C, and immersed in medium [DMEM with 4.5 g/liter glucose, supplemented with 15 mM HEPES, 110 mg/liter Na-pyruvate, 1 g/liter NaHCO3, and 0.5 g/liter NaCl (pH 7.5)], which is being pumped at a set rate through the chamber and collected in 2-min fractions. Aggregates are perifused for a period of 1.5 h to stabilize hormone release before any test substances are applied to the cells. A two-way valve system enables a smooth uninterrupted switch between a rest-stage, during which only medium (with vehicle if appropriate) is passed through the chamber, and an exposure stage, during which test substances (dissolved in the same perifusion medium) perifuse the aggregates. Fractions (0.5 ml per 2 min) are collected in 100 µl 2% BSA (Serva) in 0.01 M phosphosaline, vortexed, and stored at –20 C before analysis. To avoid any time delay in the onset of action of receptor antagonists, the antagonists were added to the perifusion medium 30 min before the start of fraction collections and were also present during the perifusion of the agonist to be blocked. In the figures exposure time to agonists is indicated (horizontal bar), taking into account the dead time in the system (4 min).
RIA
Hormone levels were measured from duplicate samples by specific competitive RIAs using the rat GH kit obtained from Dr. A. F. Parlow [National Hormone and Peptide Program, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), Harbor-UCLA Medical Center, Torrance, CA] with monkey antirat GH (G-5) antiserum and rat GH-RP-2 as reference preparation. The reliability of the assays has been shown in previous work (44, 47). Both intra and interassay variations were less than 10%. Labeling of GH with I125 was done with the Iodogen method according to the manufacturer’s instructions (Pierce, Perbio Sciences, Erembodegem, Belgium). Values were expressed in nanogram equivalents of NIDDK rat GH-releasing peptide-2 standard. The average basal GH release was 14 ng per 2 min in untreated and T3-treated aggregates, 12 ng per 2 min in Dex-treated aggregates, and 11 ng per 2 min in E2-treated aggregates.
cAMP measurements
Aggregates were cultured as mentioned above. Cells were subsequently placed in the same medium but with 0.5 mM IBMX and placed in nontreated 35-mm-diameter petri dishes at a density of roughly 1 x 106 cells/dish. The cells were then incubated in culture medium with IBMX and 5-HTR agonists or vehicle for 30 min at 37 C and in a CO2-humidified incubator after which aggregates were collected and cell lysis was obtained through the addition of 0.1 M HCl and sonication. cAMP content was quantified through enzymatic immunoassay kit from Assay Designs (Ann Arbor, MI) (48).
Data expression and statistics
All data from the perifusion experiments are expressed as percentage of basal hormone secretion defined as the average secretion before each drug application. Values are expressed as mean ± SEM. Statistical analysis was performed using either a one-way ANOVA with Fisher’s least significant differences (LSD) multiple comparison test on the cumulative secretion for the duration of drug application [area under the curve (AUC)] or two-way ANOVA with Fisher’s LSD multiple comparison test using individual values of each 2-min fraction. All data from the cAMP experiments are expressed as percentage of control cAMP production and values are expressed as mean ± SEM. One-way ANOVA with Fisher’s LSD multicomparison test was used for the statistical analysis after data were log transformed due to heterogeneity of variance.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, 5-HT readily increased GH secretion (P < 0.005, P < 0.02, and P < 0.001 for 10 nM, 100 nM, and 1 µM, respectively), but there was no dose response at the doses used (AUC = 59, 52, and 57% above basal release, respectively). The GH responses to 5-HT (10 and 100 nM and 1 µM) in aggregates cultured in hormone-free medium was lower in magnitude than that seen in the E2 (1 nM) condition (P < 0.001 at all doses), with significant increases at all concentrations (P < 0.02, P < 0.005, and P < 0.01) (Fig. 1B). The response appeared already maximal at the 10 nM dose.
|
). The AUC at 1 nM 5-HT was not significantly different from that at 3 nM 5-HT, but there was a significant difference between the 1 and 10 nM dose (P < 0.05).
The presence of thyroid hormone and glucocorticoids during culture was also tested because these hormones also can influence GH release in response to various peptides and monoamines (49, 50, 51). The GH response was not markedly altered when aggregates were cultured in 10 nM Dex (P < 0.01, P < 0.001, and P < 0.01 for 10 and 100 nM and 1 µM 5-HT) or 1 nM T3 (P < 0.02, P < 0.001, and P < 0.001 for 10 and 100 nM and 1 µM) (Fig. 1D).
For all further experiments presented below, aggregates were cultured with the 1 nM E2 supplement unless otherwise stated, and the routine dose of 5-HT used was 10 nM.
Nonselective 5-HTR antagonists block the effect of 5-HT on GH release
To reach full blockade of the 5-HTRs, antagonists were used at a concentration at least 100 times larger than the inhibitory constant (Ki) value of their receptor binding at the respective 5-HTRs (estimated from competition with agonist tracer). The dose of 5-HT was fixed at 10 nM (Ki of 5-HT for agonist binding ranges between 
1 and 10 nM for most 5-HTRs; see http://pdsp.med.unc.edu/kidb.php).
Aggregates perifused prior and during exposure to 5-HT with 1 µM of the nonselective 5-HTR antagonists methysergide or methiothepin (blocking 5-HTR1, -2, -5, -6, and -7) (52, 53, 54, 55, 56, 57) displayed a significantly reduced GH response, compared with 5-HT alone (P < 0.001 and P < 0.001, respectively) (Fig. 2, A and B). When the two antagonists were combined, the GH response to 5-HT was almost completely abolished (Fig. 2C). The application of methysergide (1 µM) on its own had no effect on basal GH release (Fig. 2D), excluding an agonistic activity by methysergide, although partial agonist activity has been previously described in some systems (58).
|
). The 5-HTR2A antagonist ketanserin (1 µM) (54, 64, 65, 66) caused the GH response to 5-HT to increase in magnitude (P < 0.001) (Fig. 3B). The 5-HTR2B/2C antagonist SB-206553 (1 µM) (61, 67, 68) significantly attenuated the GH response (64.5 ± 7% decrease) (P < 0.005), whereas the 5-HTR2C antagonist RS-102221 (1 µM) (52, 53, 61, 68) was without effect (Fig. 3, B and C). The 5-HTR3 antagonist ondansetron (1 µM, n = 3) (69, 70), the 5-HTR4 antagonist GR113808 (1 µM, n = 3) (71, 72, 73), and the 5-HTR6 antagonist Ro 04–6790 (1 µM, n = 3) (74) were inactive (data not shown). Finally, the 5-HTR7/5A antagonist SB-269970 (1 µM) (75) caused a 47.8 ± 17% decrease of the response (Fig. 3D) (P < 0.05).
|
) (P < 0.01). Surprisingly, this was also the case when the three selective 5-HTR2 antagonists (ketanserin, SB-206553, and RS-102221) (each 1 µM) were perifused together (P = 0.000037, compared with the release still detectable with SB-206553 alone), even though ketanserin and RS-102221 alone were ineffective in decreasing the response and the 5-HTR2B antagonist did not fully block the response (Fig. 4B). Complete annihilation of the GH response was also seen with a combination of ketanserin and SB-206553 (Fig. 4C) (P < 0.05, compared with the release still detectable with SB-206553 alone), but no change in GH response was seen with a combination of ketanserin and RS-102221 (data not shown, n = 3).
|
). DOI, a partial agonist of the 5-HTR2A/2C and somewhat weaker than the 5-HTR2B (54, 76, 77) (Fig. 5A), caused an increase in GH release at both 100 nM and 1 µM (P < 0.005) as did mCpp, a partial agonist at the 5-HTR2B/2C (72, 76, 78) (P < 0.01 and P < 0.05 for 100 nM and 1 µM, respectively) (Fig. 5B). The magnitude of the response was markedly lower than that evoked by 5-HT tested concurrently (P < 0.005 at 100 nM). The 5-HTR2A/2B/2C agonist -methyl 5-HT (100 nM and 1 µM) (54, 65, 66, 76, 79) caused a significant increase in GH release (P < 0.03 and P < 0.04 for 100 nM and 1 µM, respectively) (Fig. 5C), the maximal effect already seen at 100 nM but of lower magnitude than that elicited by 5-HT (P < 0.002 at 100 nM). Tryptamine, a medium-potent agonist of the 5-HTR7 and low potency agonist of 5-HTR2B (54, 57), significantly increased GH release at 10 and 100 nM (P < 0.04 and P < 0.01, respectively) (Fig. 5D), although no dose response was seen and efficacy remained below that of 5-HT (P < 0.00001 at 10 nM). A strong (P < 0.01 at 10 nM, P < 0.001 at 100 nM) and dose-dependent GH response (P < 0.05 between 10 and 100 nM) was seen with 5-carboxytryptamine (5-CT), an agonist of 5-HTR1A/B/D (62, 65, 80); 5-HTR5 (54, 55, 81); and 5-HTR7 (57, 82) (Fig. 5E). The magnitude of response remained below that to 5-HT (P < 0.03 at 10 nM; P < 0.02 at 100 nM). The 5-HTR1A agonist 8-OHDPAT (65, 83) (Fig. 5F) and sumatriptan (data not shown, n = 3), an agonist at the 5-HTR1B/D (54, 84), did not affect GH release. When comparing the efficacies (AUC found at 100 nM or 1 µM of each agonist divided by that of the 10 nM dose of 5-HT) of each agonist used, 5-CT had the highest efficacy (0.82), followed by -methyl 5-HT (0.58) greater than tryptamine (0.48) equal to or greater than DOI (0.43) greater than mCPP (0.37).
|
20 nM) (82) without affinity at the 5-HTR5A (81, 90), 5-HTR1D (91), and 5-HTR2B (52, 54) and can therefore be of further help in discriminating between the involvement of the 5-HTR2A or 5-HTR1D and between the 5-HTR5A and 5-HTR7.
At a dose of 1 µM, spiperone diminished the GH response to 5-HT only very slightly, although a 3 µM dose considerably attenuated the response (P < 0.005; P < 0.01 vs. basal release) (Fig. 6A). When combined with the 5-HTR2B/2C antagonist SB-206553 (1 µM), spiperone caused a partial inhibition of the GH response when used at 1 µM (P < 0.01 vs. 5-HT, P < 0.01 vs. basal release) (Fig. 6B), consistent with the above conclusion that it is not the 2A/2B blockade that is responsible for fully annihilating the GH response to 5-HT and the possible participation of the 5-HTR5A in establishing the GH response because the 5-HTR5A is not blocked by spiperone. Spiperone completely abolished the response when a 3 µM dose was used together with the 5-HTR2B antagonist SB-206553 (Fig. 6B). At this concentration spiperone fully blocks the 5-HTR7, explaining the full blockade when coadministered with the 5-HTR2B/2C antagonist SB-206553.
|
0.3 nM) but little the 5-HTR2B (54, 92) (Ki 150 nM). To further dissect out the putative participation of the former receptors, we tested the blocking behavior of the 5-HTR7/5 antagonist SB-269970 with 5-CT as agonist, with the expectation that SB-269970 would be more effective in reducing the GH response than with 5-HT as agonist. A 10 nM dose of 5-CT stimulated GH release as expected, but SB-269970 (1 µM) attenuated the GH response (P < 0.04) to a comparable extent that it did on the GH response to 10 nM 5-HT (39.6 ± 25% inhibition with 5-CT vs. 47.8 ± 17% with 5-HT) (data not shown, n = 3). These data are consistent with the contention that the GH response to 5-CT is not mediated solely by the 5-HTR7 but also indicate that at 10 nM 5-CT already acts at other 5-HTR(s) (such as the 5-HTR2B and/or 5-HTR1B).
Characterization of the GH response to 5-HT in aggregates cultured in hormone-free medium
For several reasons it looked important to also test the involvement of 5-HTRs in aggregates cultured in hormone-free medium. These reasons included: 1) the expression of different 5-HTRs is dependent or modulated by estrogen; 2) because the 5-HTR2C is repressed by estrogen, its participation in the GH response to 5-HT in the above experiments was considered negligible, and hence, the 5-HTR2B could be studied with the 5-HTR2B/2C antagonist SB-206553, but in the absence of estrogen a more selective 5-HTR2B antagonist and a selective 5-HTR2B agonist needed to be used to ascertain whether only the 5-HTR2B or also the 5-HTR2C mediates the GH response; and 3) in the above experiments, we studied the involvement of the 5-HTR1B only by using a 5-HTR1B/D antagonist. Because the expression of the 5-HTR1B is higher in the absence of estrogen than in its presence, it looked necessary to reexamine it again with a selective 5-HTR1B agonist and antagonist.
As shown in Fig. 7A, in aggregates cultured in hormone-free medium, the 5-HTR7/5 antagonist still diminished the GH response (–21.8%) (P < 0.01), whereas the 5-HTR2B/2C antagonist was strongly depressing it (86.1%), as expected (P < 0.01 vs. 5-HT; P < 0.001 vs. basal release). Interestingly, the impact of SB-269970 (–21.8%) was only half that in E2-supplemented culture (–41%; see Fig. 3).
|
), whereas the selective 5-HTR2B agonist BW 723C86 (100 nM) (93) significantly (P < 0.001) increased basal GH secretion (Fig. 7D). The GH response to 5-HT was slightly but not significantly (P = 0.07) affected by RS-102221 (Fig. 7B), suggesting that the 5-HTR2C does not significantly contribute to the response by 5-HT in hormone-free aggregates.
The selective 5-HTR1B antagonist SB-224289 (1 µM) slightly diminished the GH response to 5-HT (P < 0.05) (Fig. 7C), similarly to that seen in E2-treated aggregates (see above), whereas the selective 5-HTR1B agonist CP 93129 (94) slightly yet significantly (P < 0.001) stimulated GH release (Fig. 7D).
5-CT stimulates cAMP production in aggregates cultured in E2-free condition
In a previous paper reporting the PRL secretory response of the aggregates used in the present study, we showed that 5-HT increased intracellular cAMP levels (1). E2 supplementation during culture significantly potentiated this response as a result of E2 induction of 5-HTR4 expression. We here report that in E2-free condition, 5-CT (100 nM), in which it cannot act via the 5-HTR4, 5-HTR5, and 5-HTR6 (not expressed), caused a 2.5-fold rise in cAMP accumulation (P < 0.001) (Fig. 8), consistent with an action at the 5-HTR7, 5-HTR2B, and 5-HTR1B.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An important issue of the present study was to determine which 5-HTR(s) mediate(s) the GH response to 5-HT. A rough indication for 5-HTR mediation is the finding that a 1 µM dose of the nonselective 5-HTR antagonists methiothepin and methysergide diminished the GH response to 10 nM 5-HT. Methiothepin is a high-affinity blocker (Ki 
10 nM) of the 5-HTR2A/B/C, 5-HTR6, 5-HTR7, and 5-HTR1D, whereas methysergide is a high-affinity blocker of the 5-HTR2A/B/C and a medium-affinity antagonist of the 5-HTR7 and 5-HTR1D/F (52, 53, 54, 55). Both antagonists display low affinity at other receptors and no blocking action at the 5-HTR3 and 5-HTR4 (98, 99). Because methiothepin and methysergide did fully block the GH response, when applied in combination, it is likely that the 5-HTR2A/B/C, 5-HTR5A/B, 5-HTR7, and some of the 5-HTR1s, but not the 5-HTR3 and 5-HTR4, are the candidate 5-HTRs that can mediate the GH response to 5-HT.
To discriminate between the candidate receptors, partially or fully selective antagonists and agonists were tested. The GH response to 10 nM 5-HT was not affected by the following antagonists of the following receptors: ondansetron (5-HTR3), GR-113808 (5-HTR4), Ro 046790 (5-HTR6), BRL 15572 (5-HTR1A/D), WAY100635 (5-HTR1A), and RS-102221 (5-HTR2C). The 5-HTR1B blocker SB-224289 displayed a small inhibitory effect on the GH response. The 5-HTR2B/2C antagonist SB-206553 behaved as an effective blocker. Participation of the 5-HTR2C was not expected in the presence of estrogen because the receptor is down-regulated. In the absence of estrogen, the 5-HTR2C is probably also not involved as the fully selective 5-HTR2B antagonist, SB-204741 (61), inhibited the GH response equally well as SB-206553, and the selective 5-HTR2C antagonist RS-102221 did not significantly reduce the GH response to 5-HT. Remarkably, ketanserin, the prototype 5-HTR2A antagonist but also a fairly potent 5-HTR1D antagonist [Ki 10 nM in the rat (85, 86)], increased the response. This apparent disinhibition is not expected to be located at the 5-HTR2A because the canonical 5-HTR2A-mediated pathway is stimulatory, although inhibitory responses have also been reported (100, 101, 102). An alternative explanation of the ketanserin effect is disinhibition of 5-HT through the 5-HTR1D, consistent with the negative coupling of 5-HTR1s to adenylyl cyclase (103, 104). SB-269970, a preferential 5-HTR7 antagonist but also displaying fairly good affinity for the 5-HTR5A (Ki 15 nM), also partially antagonized the GH response to 5-HT. Taken together, these data suggest that the GH response to 5-HT is mediated by the 5-HTR2B, the 5-HTR7, and/or 5-HTR5A, with probably a contribution of the 5-HTR1B. In addition, the results obtained with ketanserin suggest that 5-HT may also have an inhibitory influence on GH secretion through either the 5-HTR2A or 5-HTR1D.
Further arguments for mediation by the 5-HTR2B, 5-HTR7, and 5-HTR1B was obtained by studies with agonists. The involvement of the 5-HTR2B was proved by the finding that the fully selective 5-HTR2B agonist BW723C86 (93) readily increased basal GH release. The stimulatory action of DOI, -methyl-5-HT, and mCpp, which are agonists at all 5-HTR2 subtypes (Ki 1–10 nM), without or with only low affinity for other 5-HTRs (54, 65, 66, 72, 76, 78), was consistent with a 5-HTR2 mediation. The lower efficacy of DOI and the even lower efficacy of mCPP are consistent with their reported partial agonism (105). The contribution of the 5-HTR1B was confirmed as the selective 5-HTR1B agonist CP 93129 (94) slightly stimulated GH release. Participation of the 5-HTR7 in GH secretion could not be tested with a fully selective agonist. However, GH release was stimulated by 5-CT, an agonist displaying highest affinity for the 5-HTR7 (Ki 1 nM) (57, 82). Because 5-CT also has affinity for the 5-HTR5A and -B (Ki 1 nM) (54, 81) and the 5-HTR1A, -B, and -D (Ki 0.3–10 nM) (54, 84) but low affinity for the 5-HTR2B (Ki 150 nM) (54, 76, 106), the effect of 5-CT at 10 nM is likely mediated mainly by the 5-HTR7 and 5-HTR1B/1D. At the 100 nM dose, 5-CT may also act through the 5-HTR2B. The GH response to 10 nM tryptamine is probably also through the 5-HTR7 because at that dose tryptamine is expected to occupy only the 5-HTR7 [Ki 15 nM at 5-HTR7 (54, 57) and 100 nM at 5-HTR2B (54, 76)]. Tryptamine, however, has also some affinity for the 5-HTR1D (Ki 40–80 nM) (107), through which it may dampen the response, a hypothesis consistent with the absence of a dose response in the effect of tryptamine.
Our data are also intriguing because the different 5-HTRs that appear to be involved do not seem to simply add linearly to the GH response. Blocking both the 5-HTR2B and 5-HTR7 or -5A simultaneously are sufficient to completely abolish the GH response. The 5-HTR2B appears to contribute most. Moreover, the GH response was also completely abolished by combined application of ketanserin and SB-206553, even though ketanserin alone enhanced the magnitude of the GH response. Thus, there seems to be a cross talk among the 5-HTR subtypes or between the cells expressing these receptors. One possible model that may explain these observations is that to have a full GH response to 5-HT, both transduction via a Gs and a Gq-coupled mechanism must occur or that actions via Gs are gated by the Gq-induced intracellular messengers. The 5-HTR2 subtypes are known to signal through Gq (108), the 5-HTR7 through Gs (82), and the 5-HTR1 through Gi/o or Gs (109). According to this model, no or little GH response will then be seen in case there is either combined 5-HTR1D and 5-HTR2B blockade or combined 5-HTR2B and 5-HTR7 blockade.
Some further support for the involvement of the 5-HTR1D can be found in the spiperone experiments. Application of the 5-HTR2B antagonist SB-206553 together with spiperone, an antagonist of the 5-HTR2A with equal potency (1 nM) and similar structure as ketanserin (54) but with no affinity for the 5-HTR1D (54, 84), did not fully block the GH response at a dose of 1 µM, whereas ketanserin + SB-206553 did. In fact, if spiperone and ketanserin would have acted through the 5-HTR2A, the spiperone + SB-206553 combination was expected to block the GH response more effectively than the ketanserin + SB-206553 combination because, unlike ketanserin, spiperone also displays blocking affinity for the 5-HTR7 (Ki 20 nM) (82), and combined blockade of 5-HTR2B and 5-HTR7 fully abolishes the GH response. At a higher dose (3 µM), spiperone did fully block the GH response when combined with SB-206553, consistent with a blocking action of spiperone at the 5-HTR7. The requirement of a higher dose is consistent with the fact that spiperone has an affinity about 10 times lower than SB-269970 for the 5-HTR7.
With respect to the putative negative regulation by the 5-HTR1D, the puzzling finding has to be dealt with that sumatriptan, an agonist at the 5-HTR1B/D/F (54, 84), did not affect GH release, even at a 1 µM dose. It is possible that sumatriptan simultaneously exerts an inhibitory actions via the 5-HTR1D and a stimulatory one via the 5-HTR1B, resulting in no observable changes in GH secretion. However, at first look there is no argument for this because BRL15572, blocking the 5-HTR1A/D (60), also failed to affect the GH response to 5-HT. The ineffectiveness of BRL15572 may be the consequence of the fact that it is also a medium-potent blocker of the 5-HTR2B (60). Consequently, inhibition of GH secretion by 5-HT via the 5-HTR1D might be compensated for by blocking the stimulatory action via the 5-HTR2B. Future experiments exploring the separate effect of selective 5-HTR1D antagonists and agonists on basal and stimulated GH release are required to further evaluate 5-HTR1-mediated negative regulation.
We were not able to gather conclusive evidence as to the participation of the 5-HTR5 in the GH response because the only selective 5-HTR5 antagonist available is potently active at the human 5-HTR5A (87) but not at the rat 5-HTR5A (88) (no data available for 5-HTR5B). Therefore we tested the participation of the 5-HTR5A by indirect approaches. First, it was found that spiperone, known to block the 5-HTR7 (Ki 20 nM) (82) but not the 5-HTR5A (Ki >1 µM) (81, 90), displayed a small inhibitory action on the GH response at a dose of 1 µM and a pronounced depression at 3 µM, indicating that the fraction of the GH response blockable by SB-269970 is not mediated by the 5-HTR5A alone but at least in addition by the 5-HTR7. The effect at 1 µM was less than that of the 5-HTR7 antagonist SB-269970, possibly because spiperone is about 10 times less potent at the 5-HTR7 than SB-269970 (Ki 1 nM). In a second approach, the 5-HTR7 blocking capacity of SB-269970 was tested in aggregates cultured in E2-free medium. Because in the latter condition, the 5-HTR5A and -B is not expressed to a detectable level (1), it was expected that SB-269970 would be ineffective on the GH response if the 5-HTR5A (or -B) was the only mediating receptor. However, we found that SB-269970 still inhibited the GH response, suggesting that the 5-HTR7 is indeed involved. Nevertheless, the GH response was less affected by SB-269970 in E2-free condition than in the E2-supplemented condition, consistent with at least a possibility for participation of the 5-HTR5A or -B in the latter. We cannot, however, exclude the possibility that treatment with E2 does in fact increase the functional output of the 5-HTR7, in addition to rendering the 5-HTR5 functional.
No studies have examined the role of 5-HTRs other than 5-HTR1B/D, 5-HTR2A/C, and 5-HTR3 subtypes in GH regulation in vivo. The present study shows for the first time that the effect of 5-HT at the pituitary level is not mediated by the 5-HTR2A, 5-HTR2C, or 5-HTR3 but by the 5-HTR2B, 5-HTR7, and 5-HTR1B. Partial effect by the estrogen-induced 5-HTR5 remains a possibility, extending our previous observation of estrogen-induced 5-HTR4 in PRL release (1). The present findings open a new avenue in understanding the complexity of 5-HT in GH regulation. It is conceivable that 5-HT participates in holding basal GH release to a certain level beyond the contribution of the GHRH/somatostatin balance. It is interesting in this context to consider a role of estrogen induction of 5-HTR4, 5-HTR5, and 5-HTR6 (1) because estrogens have been shown to increase not only basal PRL release but also basal GH release by a pituitary site of action (110). We previously reported that the 5-HTR4 is up-regulated by E2 treatment and is responsible in mediating the effect of 5-HT on the PRL secretion of anterior pituitary cells (1). What is of particular interest is that the 5-HTRs mediating the GH response are different from the one involved in mediating the PRL response, suggesting that there is a necessity to keep these two systems separate during adaptive functional changes in the pituitary. The present studies demonstrate a unique and complicated role for 5-HT at the level of the pituitary in the regulation of hormone secretion that had not previously been demonstrated.
In conclusion, the present data show a stimulatory action at the pituitary level of 5-HT on GH release and that this effect is mediated by several 5-HTRs. The largest part of the response is mediated by the 5-HTR2B with a contribution of the 5-HTR7 and the 5-HTR1B. The 5-HTR1D may also be involved in a negative fashion. These receptors appear to cross talk, which may be important for stabilizing basal GH release at the pituitary level.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The authors have nothing to disclose.
First Published Online June 21, 2007
Abbreviations: AUC, Area under the curve; 5-CT, 5-carboxytryptamine; Dex, dexamethasone; DOI, 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane; E2, estradiol; 5-HT, 5-hydroxytryptamine; 5-HTR, 5-HT receptor; IBMX, isobutylmethylxantine; Ki, inhibitory constant; LSD, least significant differences; mCpp, m-chlorophenyl piperazine; 8-OHDPAT, 7-(dipropylamino)tetralin-1-ol-8-hydroxy-2-(di-n-propylamino)tetralin; PRL, prolactin; VIP, vasoactive intestinal peptide.
Received January 10, 2007.
Accepted for publication June 11, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA)-stimulated GH secretion in prepubertal male rats. Eur J Endocrinol 144:535–541[Abstract]-subunit genes by gonadotropin-releasing hormone and estradiol-17ß in single rat pituitary cells during aggregate cell culture. Endocrinology 144:388–399-dihydrotestosterone, and dexamethasone. Endocrinology 126:2257–2266[Abstract]
