Does a Melatonin‐Dependent Circadian Oscillator in the Pars Tuberalis Drive Prolactin Seasonal Rhythmicity?

The pars tuberalis (PT) of the adenohypophysis expresses a high density of melatonin receptors and is thought to be a crucial relay for the actions of melatonin on seasonal rhythmicity of prolactin secretion by the pars distalis (PD). In common with the suprachiasmatic nucleus of the hypothalamus and most other peripheral tissues, the PT rhythmically expresses a range of 'clock genes'. Interestingly, this expression is highly dependent upon melatonin/photoperiod, with several aspects unique to the PT. These observations led to the establishment of a conceptual framework for the encoding of seasonal timing in this tissue. This review summarises current knowledge of the morphological, functional and molecular aspects of the PT and considers its role in seasonal timing. The strengths and weaknesses of current hypotheses that link melatonin action in the PT to its seasonal effect on lactotrophs of the PD are discussed and alternative working hypotheses are suggested.     

Production of a prolactin releasing factor by the ovine pars tuberalis

The anterior pituitary hormone prolactin is involved in the regulation of a wide variety of processes including mammary growth, lactation, reproduction and pelage growth (1). Correspondingly, in seasonal breeders, pronounced photoperiod-driven, seasonal changes occur in the levels of prolactin secretion (2). These effects are thought to be mediated by the pineal hormone melatonin, which acts as a humoral indicator of photoperiod (3). Melatonin is thought to act through hypothalamic sites to control the gonadotrophic axis (4–6), but the sites through which melatonin modulates prolactin remain to be established. One possibility is that melatonin acts at the level of the hypothalamus to modulate the release of the hypothalamic prolactin inhibitory factor, dopamine (7). However, recent evidence from hypothalamo-pituitary disconnection experiments performed in the ram suggests that the photoperiodic modulation of prolactin secretion can occur independently of the hypothalamus, presumably due to direct effects of melatonin on the anterior pituitary (8). The only identified site of melatonin receptor expression within the ovine pituitary is the pars tuberalis (PT) (9), the cells of which are almost completely non-lactotrophic (10). Thus the possibility exists that factors released by the PT regulate the activity of lactotrophs in the pars distalis (PD). We have investigated this hypothesis using a range of co-culture and medium-conditioning experiments on primary cultures of ovine PT and PD cells, and here we report that PT  cells secrete an unidentified factor that is a potent stimulus of prolactin secretion by PD cells.     

Melatonin Receptors in the Brain and Pituitary Gland of Hypothalamo-Pituitary Disconnected Soay Rams

In Soay rams in which the pituitary gland has been surgically separated from the hypothalamus, blood prolactin concentrations vary in response to changes in photoperiod and the administration of melatonin, as in intact animals, providing evidence that melatonin acts within the pituitary gland to control prolactin secretion. In this study the presence of potentially functional melatonin receptors in the pars tuberalis and zona tuberalis (PT/ZT) of hypothalamo-pituitary disconnected (HPD) Soay rams is confirmed using both in vitro autoradiography with the ligand 2-(¹²⁵I)- iodomelatonin and in situ hybridization for the melatonin receptor. There was no effect of the HPD operation on the pattern and quantity of 2-(¹²⁵I)iodomelatonin binding in the brain demonstrating that this binding is independent of hypothalamic regulation. The possibility that melatonin may control prolactin secretion directly via specific receptors on lactotrophs was investigated using dual in situ hybridization with a (³⁵S) labelled probe for the ovine melatonin receptor (Mel1a_b) and a Digoxigenin labelled probe for ovine prolactin. Melatonin receptor gene expression was observed in the PT/ZT in both intact and HPD rams, however, there was no colocalization with prolactin gene expression; only in the ZT was there a close association between cells expressing the melatonin receptor and lactotrophs. The results provide strong support for the view that melatonin acts via the PT/ZT to mediate the effects of photoperiod on the seasonal cycle in prolactin secretion.     

Rhythms in clock proteins in the mouse pars tuberalis depend on MT1 melatonin receptor signalling

Melatonin provides a rhythmic neuroendocrine output, driven by a central circadian clock that encodes information about phase and length of the night. In the hypophyseal pars tuberalis (PT), melatonin is crucial for rhythmic expression of the clock genes mPer1 and mCry1, and melatonin acting in the PT influences prolactin secretion from the pars distalis. To examine further the possibility of a circadian clockwork functioning in the PT, and the impact of melatonin on this tissue, we assessed circadian clock proteins by immunohistochemistry and compared the diurnal expression in the PT of wild type (WT), and MT1 melatonin receptor-deficient (MT1-/-) mice. While in the PT of WT mice mPER1, mPER2, and mCRY1 showed a pronounced rhythm, mCRY2, CLOCK, and BMAL1 were constitutively present. Despite reported differences in maximal levels and timing of mCry1, mPer1, and mPer2 RNAs, the corresponding protein levels peaked simultaneously during late day, suggesting a codependency for their stabilization and/or nuclear entry. MT1-/- mice had reduced levels of mPER1, mCRY1, CLOCK and BMAL1, consistent with the earlier reported reduction in mRNA expression of these clock genes. Surprisingly, mPER2-immunoreaction was constitutively low, although mPer2 was rhythmically expressed in the PT of MT1-/- mice. This suggests that mPER2 is degraded due to the reduced levels of its stabilizing interaction partners mPER1 and mCRY1. The results show that melatonin, acting through the MT1, determines availability of the circadian proteins mPER1, mPER2 and mCRY1 and thus plays a crucial role in regulating rhythmicity in PT cells.     

RAPID COMMUNICATION oPer1 is an Early Response Gene Under Photoperiodic Regulation in the Ovine Pars Tuberalis

Mammalian Per1 (or RIGUI ) is a recently described putative clock gene that is expressed in the suprachiasmatic nucleus. It is also expressed in the pars tuberalis (PT) of the pituitary, where melatonin appears to drive its expression. This study examines the regulation of Per1 expression. In ovine PT cells, oPer1 is an early response gene transiently expressed after stimulation with forskolin, but melatonin has no independent effect on its expression. In sheep, PT tissue photoperiodic background influences the magnitude or timing of expression of oPer1 2  h after lights-on. These data demonstrate that oPer1 mRNA is elevated in the PT following the decline in night-time melatonin, and that the amplitude or timing of this elevation is dependent upon the duration of the nocturnal melatonin signal.     

MT1 melatonin receptor mRNA expressing cells in the pars tuberalis of the European hamster: effect of photoperiod

Melatonin, secreted only during the night by the pineal gland, transduces the photoperiodic message to the organism. One important target for the hormone is the pars tuberalis (PT) of the adenohypophysis which displays a very high number of melatonin binding sites in mammals and is implicated in the seasonal regulation of prolactin secretion. To gain insight into the mechanism by which the melatonin signal is decoded in the PT, we studied the effect of photoperiod on the PT cells expressing the MT1 melatonin receptor in a highly photoperiodic species, the European hamster. Recently, we showed that, in the rat, the MT1 receptor mRNA is expressed in PT-specific cells characterized by their expression of beta-thyroid stimulating hormone (beta-TSH) along with the alpha-glycoprotein subunit (alpha-GSU). As the cellular composition of the PT shows variability among species, we first identified the cell type expressing the MT1 receptor in the European hamster by combining immunocytochemistry and nonradioactive in situ hybridization for the MT1 receptor mRNA. Our results show that, in the European hamster, as in the rat, the MT1 receptor is only expressed by the PT-specific-cells, beta-TSH and alpha-GSU positive. In a second step, we analysed the effects of photoperiod on the MT1 mRNA, and on beta-TSH and alpha-GSU both at the mRNA and protein levels. Our data show that, compared to long photoperiod, short photoperiod induces a dramatic decrease of MT1, beta-TSH and alpha-GSU expression. Protein levels of beta-TSH and alpha-GSU were also dramatically reduced in short photoperiod. Together, our data suggest that melatonin exerts its seasonal effects in the PT by signalling to PT specific-cells through the MT1 receptor subtype.     

Melatonin regulates the synthesis and secretion of several proteins by pars tuberalis cells of the ovine pituitary

The pars tuberalis (PT) of the pituitary may be an important target for melatonin action, but the secretory output of the melatonin-responsive cells is unknown. Using [(35) S]methionine, protein synthesis and secretion have been studied in primary cultures of ovine PT cells, and analysed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Only 4% of the labelled proteins appeared in the medium with the majority retained in the cells. Stimulation of the cells with 10μM forskolin increased the accumulation of several labelled proteins in the medium without corresponding changes in the cell (72, 62, 44, 39, 29, 24, 23, 18 and 14 kd). Two-dimensional gel electrophoresis showed the proteins to have mildly acidic isoelectric points. Melatonin (1 μM) counteracted the stimulatory effect of forskolin on all but one (23 kd) of these secreted proteins. Immunoprecipitation showed this to be prolactin. Furthermore, melatonin alone appeared to have an inhibitory effect on the synthesis and release of proteins into the medium. The synthesis and secretion of the melatonin-responsive proteins was not inhibited by actinomycin D (1 μg/ml), indicating control at the translational level. This contrasts with the regulation of prolactin which is actinomycin D-sensitive. Pulse-chase experiments demonstrated that it requires 30 min for the secretory proteins to appear in the medium, consistent with intracellular processing and packaging prior to secretion. The secretory proteins labelled in the ovine PT, and responsive to melatonin, did not appear to be specific to the PT, as a similar profile of labelled secretory proteins was produced in primary cultures of pars distalis cells. However, melatonin had no effect on the synthesis and secretion of proteins by the pars distalis. These results demonstrate that in the ovine PT melatonin regulates the synthesis and export of several secretory proteins. These are possibly packaging proteins of secretory granules, similar to the granin family of proteins. Thus, the results confirm that melatonin-responsive cells are secretory cells and further imply that the PT-specific product is not a protein.     

Daily variations in melatonin receptor density of rat pars tuberalis and suprachiasmatic nuclei are distinctly regulated

Suprachiasmatic nuclei (SCN) and pars tuberalis (PT) are two structures in the rat exhibiting high affinity receptors for melatonin. Melatonin receptor density in these two structures was previously shown to be inversely related to endogenous ligand concentration, thus elevated at daytime. We now demonstrate that, in the PT, these daily variations are directly induced by the circadian rhythm of plasma melatonin concentration. Variations persist in constant darkness and can only be blocked by pinealectomy. Thus, autoregulation loop of melatonin receptors determines the circadian rhythm in PT melatonin receptor de density. However, this process of densensitization does not determine the daily variations in SCN melatonin receptor density. Indeed, in the SCN, the light/dark cycle is the regulatory factor: melatonin receptor density was shown to be specifically reduced during the night even in pinealectomized animals, while one h light was shown to reverse this nocturnal decrease in the SCN. Moreover, this darkness-induced down-regulation of SCN melatonin receptors has a masking effect on the earlier shown ligand-dependent desensitization process in this structure. This explain why, in constant darkness, SCN melatonin receptor density did not show any variation throughout the 24 h subjective day and night, although the circadian rhythm of melatonin persisted. These results thus clearly show that although daily rhythms in the density of melatonin receptors are identical in the PT and in the SCN, their regulation is totally different in each of these two structures.     

Photic regulation of mt1 melatonin receptors in the Siberian hamster pars tuberalis and suprachiasmatic nuclei: involvement of the circadian clock and intergeniculate leaflet

In the Siberian hamster suprachiasmatic nuclei and pars tuberalis of the pituitary, high affinity mt1 melatonin receptors are present. We have previously shown that night applied light pulse induced an increase in mt1 mRNA expression in the suprachiasmatic nuclei of this species, independently of the endogenous melatonin. Here, we report the photic regulation of melatonin receptor density and mRNA expression in the suprachiasmatic nuclei and pars tuberalis of pinealectomized Siberian hamsters and the implication in this control of either the circadian clock or the intergeniculate leaflet. The results show that: (1) A 1-h light pulse, delivered during the night, induces a transitory increase in mt1 mRNA expression in the suprachiasmatic nuclei and pars tuberalis. After 3 h this increase has totally disappeared (suprachiasmatic nuclei) or is greatly reduced (pars tuberalis). (2) The melatonin receptor density, in the suprachiasmatic nuclei, is not affected by 1 or 3 h of light, while it is strongly increased in the pars tuberalis. (3) In hamsters kept in constant darkness, the mt1 mRNA rise is gated to the subjective night in the suprachiasmatic nuclei and pars tuberalis. In contrast, the light-induced increase in melatonin binding is also observed in the subjective day in the pars tuberalis. (4) intergeniculate leaflet lesion totally inhibits the mt1 mRNA expression rise in the suprachiasmatic nuclei, while it has no effect on the light-induced increase in mt1 mRNA in the pars tuberalis. However, the light-induced increase in melatonin receptor density is totally prevented by the intergeniculate leaflet lesion in the pars tuberalis. These results show that: (1) the photic regulations of mt1 mRNA expression and receptor density are independent of each other in both the suprachiasmatic nuclei and pars tuberalis; and (2) the circadian clock and the intergeniculate leaflet are implicated in the photic regulation of melatonin receptors but their level of action differs totally between the suprachiasmatic nuclei and pars tuberalis.     

Seasonal regulation of melatonin receptors in rodent pars tuberalis: correlation with reproductive state

Summary Using quantitative autoradiography, we have studied the relationship between melatonin receptor density in the pars tuberalis (PT) and photoperiodic changes in sexual activity in a photoperiodic (Syrian hamster), and a non photoperiodic (rat) rodent. Syrian hamsters exposed to short photoperiod (SP) for 13 weeks or kept in long photoperiod (LP) with daily melatonin injections for 8 weeks, had both gonadal regression and a significant decrease in 2-¹²⁵I-melatonin binding site density in the PT when compared to controls. In contrast, when the animals were sexually active, photorefractory Syrian hamsters exposed to SP for 27 weeks, or rats kept for 13 weeks in SP, the PT melatonin receptor density was similar to that of control animals kept in LP. These results show clearly that a correlation exists between gonadal status and density of melatonin receptors in the PT and suggest that the PT could be the site where melatonin mediates its effects on seasonal function.     

Evidence for RGS4 modulation of melatonin and thyrotrophin signalling pathways in the pars tuberalis

In mammals, the pineal hormone melatonin is secreted nocturnally and acts in the pars tuberalis (PT) of the anterior pituitary to control seasonal neuroendocrine function. Melatonin signals through the type 1 Gi-protein coupled melatonin receptor (MT1), inhibiting adenylate cyclase (AC) activity and thereby reducing intracellular concentrations of the second messenger, cAMP. Because melatonin action ceases by the end of the night, this allows a daily rise in cAMP levels, which plays a key part in the photoperiodic response mechanism in the PT. In addition, melatonin receptor desensitisation and sensitisation of AC by melatonin itself appear to fine-tune this process. Opposing the actions of melatonin, thyroid-stimulating hormone (TSH), produced by PT cells, signals through its cognate Gs-protein coupled receptor (TSH-R), leading to increased cAMP production. This effect may contribute to increased TSH production by the PT during spring and summer, and is of considerable interest because TSH plays a pivotal role in seasonal neuroendocrine function. Because cAMP stands at the crossroads between melatonin and TSH signalling pathways, any protein modulating cAMP production has the potential to impact on photoperiodic readout. In the present study, we show that the regulator of G-protein signalling RGS4 is a melatonin-responsive gene, whose expression in the PT increases some 2.5-fold after melatonin treatment. Correspondingly, RGS4 expression is acutely sensitive to changing day length. In sheep acclimated to short days (SP, 8 h light/day), RGS4 expression increases sharply following dark onset, peaking in the middle of the night before declining to basal levels by dawn. Extending the day length to 16 h (LP) by an acute 8-h delay in lights off causes a corresponding delay in the evening rise of RGS4 expression, and the return to basal levels is delayed some 4 h into the next morning. To test the hypothesis that RGS4 expression modulates interactions between melatonin- and TSH-dependent cAMP signalling pathways, we used transient transfections of MT1, TSH-R and RGS4 in COS7 cells along with a cAMP-response element luciferase reporter (CRE-luc). RGS4 attenuated MT1-mediated inhibition of TSH-stimulated CRE-luc activation. We propose that RGS4 contributes to photoperiodic sensitivity in the morning induction of cAMP-dependent gene expression in the PT.     

Both Pertussis Toxin-Sensitive and Insensitive G-Proteins Link Melatonin Receptor to Inhibition of Adenylate Cyclase in the Ovine Pars Tuberalis

Bordetella pertussis toxin (islet activating protein, IAP) has been used to investigate the G-proteins involved in mediating the action of the melatonin receptor. Melatonin inhibits iorskolin-stimulated cyclic AMP production in ovine pars tuberalis (PT) cells. In cells treated with IAP for 16 h this response is attenuated in a dose-dependent manner, but not abolished. IAP catalyses the incorporation of [³² P-ADP]ribose into a 41 kd protein present in PT membranes, but this labelling can be reduced if PT cells are preincubated with IAP for 16 h. Treatment of crude membrane preparations with IAP (20 /ig/ml) suppresses the binding of 2-[¹²⁵ l]iodomelatonin by 20%, whereas 1 mM GTP alone reduces binding by 40%, and in combination with IAP its effect is additive (60% inhibition). Therefore, these results indicate that the melatonin receptor acts via two G-proteins, one pertussis toxin-sensitive and the other pertussis toxin-insensitive.     

Seasonal variations of clock gene expression in the suprachiasmatic nuclei and pars tuberalis of the European hamster (Cricetus cricetus)

In mammals, day length (photoperiod) is read and encoded in the main circadian clock, the suprachiasmatic nuclei (SCN). In turn, the SCN control the seasonal rhythmicity of various physiological processes, in particular the secretion pattern of the pineal hormone melatonin. This hormone then operates as an essential mediator for the control of seasonal physiological functions on some tissues, especially the pars tuberalis (PT). In the European hamster, both hormonal (melatonin) and behavioral (locomotor activity) rhythms are strongly affected by season, making this species an interesting model to investigate the impact of the seasonal variations of the environment. The direct (on SCN) and indirect (via melatonin on PT) effect of natural short and long photoperiod was investigated on the daily expression of clock genes, these being expressed in both tissues. In the SCN, photoperiod altered the expression of all clock genes studied. In short photoperiod, whereas Clock mRNA levels were reduced, Bmal1 expression became arrhythmic, probably resulting in the observed dramatic reduction in the rhythm of Avp expression. In the PT, Per1 and Rev-erbalpha expressions were anchored to dawn in both photoperiods. The daily profiles of Cry1 mRNA were not concordant with the daily variations in plasma melatonin although we confirmed that Cry1 expression is regulated by an acute melatonin injection in the hamster PT. The putative role of such seasonal-dependent changes in clock gene expression on the control of seasonal functions is discussed.     

Regulation of Melatonin Receptors in the Pars Tuberalis of Syrian Hamsters Transferred from Long to Short Photoperiod: Implication of Melatonin and Testosterone

The exposure of long day seasonal breeders to a short photoperiod (SP) induces both sexual quiescence and a decrease in pars tuberalis (PT) melatonin receptor density. Therefore, we studied the respective roles of melatonin and testosterone on the regulation of PT melatonin receptors in Syrian hamsters transferred from long photoperiod (LP) to SP. Compared with intact sexually active animals in LP, the density of melatonin receptors was not affected by the absence of melatonin after removal of the pineal gland from animals kept in either SP or LP. In contrast, the presence of a long melatonin peak in the blood which induces gonadal atrophy induced a significant decrease in binding capacity. The SP-induced decrease in PT melatonin receptor density was also observed in castrated animals showing that it was directly regulated by melatonin, independently of circulating testosterone concentrations. However, the absence of testosterone induced an increased binding in LP, while increasing blood testosterone concentration after implantation of one testosterone-filled silastic tube resulted in a decrease in binding both in LP-and in SP-animals. These results indicate that testosterone induces a photoperiod-independent decrease in PT melatonin receptor density. In summary, these results show that both melatonin and testosterone have negative regulatory effects on the density of PT melatonin receptors.     

Melatonin induces gene-specific effects on rhythmic mRNA expression in the pars tuberalis of the Siberian hamster (Phodopus sungorus)

In mammals, circadian and photoperiodic information is encoded in the pineal melatonin signal. The pars tuberalis (PT) of the pituitary is a melatonin target tissue, which transduces photoperiodic changes and drives seasonal changes in prolactin secretion from distal lactotroph cells. Measurement of photoperiodic time in the PT is believed to occur through melatonin dependent changes in clock gene expression, although it is unclear whether the PT should be considered a melatonin sensitive peripheral oscillator. We tested this hypothesis in the Siberian hamster (Phodopus sungorus) firstly by investigating the effects of melatonin injection, and secondly by determining whether temporal variation in gene expression within the PT persists in the absence of a rhythmic melatonin signal. Hamsters preconditioned to long days were treated with melatonin during the late light phase, to advance the timing of the nocturnal melatonin peak, or placed in constant light for one 24 h cycle, thereby suppressing endogenous melatonin secretion. Gene expression in the PT was measured by in situ hybridization. We show that melatonin rapidly induces cry1 mRNA expression without the need for a prolonged melatonin-free interval, acutely inhibits mt1 expression, advances the timing of peak rev-erb alpha expression and modulates per1 expression. With the exception of cry1, these genes continue to show temporal changes in expression over a first cycle in the absence of a melatonin signal. Our data are consistent with the hypothesis that the hamster PT contains a damped endogenous circadian oscillator, which requires a rhythmic melatonin signal for long-term synchronization.     

Tissue-specific abolition of Per1 expression in the pars tuberalis by pinealectomy in the Syrian hamster

Melatonin secretion by the pineal gland transduces photoperiod into a neuroendocrine signal. In the pars tuberalis (PT), we have shown that photoperiod modifies the amplitude of the clock gene Per1. The aim of this study was to test whether the endogenous melatonin signal is required for rhythmic expression of Per1 in the PT. Male Syrian hamsters housed in long days (LD, 16:8h light:dark) were pinealectomized and Per1 mRNA expression studied by in situ hybridization. Pinealectomy abolished the rhythm of Per1 expression in the PT, but had no effect on Per1 expression in the suprachiasmatic nucleus (SCN), or the ventromedial nucleus (VMH) of the hypothalamus. Interestingly, a single night-time injection of melatonin (25 microg), given to pinealectomized animals, failed to restore Per1 expression in the PT. These data demonstrate that Per1 expression in the PT is driven by melatonin, and that the features of the endogenous signal through which the Per1 expression is achieved cannot be reproduced by a single melatonin injection.     

Diurnal change of thyroid-stimulating hormone mRNA expression in the rat pars tuberalis

Thyroid-stimulating hormone (TSH)-producing cells (TSH cells), which account for a large fraction of the cells in the rat pars tuberalis (PT), have been found to express MT1 melatonin receptor and mammalian clock genes at high densities. Although these findings suggest that TSH production in the rat PT is regulated by melatonin and/or the biological clock, there have been no studies focusing on the diurnal change and regulation mechanism of TSH production in the rat PT. Therefore, in the present study, we examined diurnal changes of in TSH beta and alpha-glycoprotein subunit (alpha GSU) mRNA expression and TSH immunoreactivity (-ir) in the rat PT, and also examined the relationship between melatonin and TSH production in vivo. Both TSH beta mRNA expression and alpha GSU mRNA expression in the PT showed diurnal variations: the expression levels were lowest at the light phase [Zeitgeber time (ZT)4] and high at the dark phase (ZT12 and ZT20). TSH-ir in the PT showed the lowest level at ZT4, as was found for mRNA expression. Interestingly, TSH-ir, which was confined to the Golgi apparatus at ZT4, spread to the cytoplasm, and most of the TSH cells in the PT were uniformly immunostained in the cytoplasm at ZT20. Despite the fact that chronic administration of melatonin suppressed TSH beta and alpha GSU mRNA expression, TSH-ir in the PT was significantly enhanced. These findings results clearly show that there are diurnal changes in TSH expression and accumulation in rat PT-TSH cells and suggest that these fluctuations are regulated by melatonin.