Decoding The Melatonin Signal

The ovine pars tuberalis has become a valuable pharmacological model for studying the function of the melatonin receptor, not only as it is amenable to cell culture and biochemical experiment, but also as it is a photoperiodically relevant target tissue. Using primary cultures of this gland it has been possible to define the acute signal trans-duction characteristics of the melatonin receptor.

Reverse transcription PCR, using degenerate primers, has been used to amplify melatonin receptor sequences from mRNA extracted from ovine pars tuberalis (oPT), and only the Mella (mti) receptor sub-type has been amplified, although we have found that this exists in two allelic forms (GenBANK accession number AF045219) (1). We have also amplified, cloned and sequenced an ovine melatonin-related receptor, which does not bind melatonin, from the ovine pars tuberalis (3), yet we have found no evidence for a sheep Mellb (mt2 receptor in this tissue or from genomic DNA (1) (unpublished observations). Competitive binding studies on oPT cell membranes using Luzindole, which has a weak selectivity for the hMellb receptor sub-type relative to hMella receptor expressed in HEK293 cells (2), confirms that the binding affinity in the oPT matches that of the Mella receptor (7) (Figure 1). This suggests that the functional melatonin receptor expressed in the sheep PT is solely the Mella receptor.

In oPT melatonin acts through a pertussis toxin-sensitive G-protein to prevent and reverse forskolin stimulated cyclic AMP levels (14,15). Notably however melatonin does not affect basal cyclic AMP levels. The recombinant sheep and human Mella receptors stably expressed in L-cells and HEK293 cells respectively have similar signal transduction characteristics (1,2). These data suggest that the melatonin receptor acts via a Gi-protein to inhibit cyclic AMP. We have also shown that melatonin can inhibit PMA induced c-fos expression in a pertussis toxin sensitive manner in oPT cells (22). As this response is not mediated through cyclic AMP, it indicates that second messenger pathways other than cyclic AMP are susceptible to acute inhibition through the

Figure 1. Competitive displacement of 2-l25I-iodomelatonin binding by Luzindole (2-benzyl-N-acetyltryptamine) from human Mella and human Mellb receptors stably expressed in HEK293 cells, and comparison to displacement from melatonin receptors in oPT. Data show that Luzindole has a 10 fold selectivity for the Mellb receptor, and receptor in oPT has an affinity like Mella for Luzindole. The data are re-drawn from Conway et al. 1997 (2) and Howell and Morgan 1991 (7). (NB. 10-14 defines the binding in the absence of Luzindole).

Figure 1. Competitive displacement of 2-l25I-iodomelatonin binding by Luzindole (2-benzyl-N-acetyltryptamine) from human Mella and human Mellb receptors stably expressed in HEK293 cells, and comparison to displacement from melatonin receptors in oPT. Data show that Luzindole has a 10 fold selectivity for the Mellb receptor, and receptor in oPT has an affinity like Mella for Luzindole. The data are re-drawn from Conway et al. 1997 (2) and Howell and Morgan 1991 (7). (NB. 10-14 defines the binding in the absence of Luzindole).

melatonin receptor in oPT cells (22). Melatonin has not been shown to have any independent acute effects through several signalling pathways in oPT cells, including activation of mitogen activated protein kinase (MAPK) (6), mobilisation of calcium (16), the turnover of phospholipases C and D (11,16), activation of protein kinase C (22) and the activation of c-fos (22). All these data suggest that the main function of melatonin is to prevent cellular activation by another external stimulus to the oPT cells, thereby rendering the cells functionally silent.

A prediction, if the above were true, would be that the activity of the PT would be quiescent during the hours of darkness co-incident with high levels of melatonin, and activated following light onset and the decline in melatonin. Therefore in the context of a photoperiodic role by melatonin, where the duration of the nocturnal mela-tonin signal determines the nature of the biological response, a simplistic interpretation would be that the PT is stimulated by another exogenous stimulus at all times when melatonin is not around. Thus under long days elevated nocturnal melatonin levels would be short, and the PT would be stimulated for an extended period. By contrast under short days the long duration of the nocturnal melatonin signal would allow the PT to be stimulated for a shorter period. This is a simplistic interpretation that takes no account of any periodicity in the availability of a humoral stimulus to the PT or of any change in receptor sensitivity to the stimulus during prolonged periods of stimulation.

In a recent paper by Sun et al., the first cloning of a mammalian homologue of Drosophila period gene, Per1 (RIGUI), was reported (23). Suprisingly, in addition to the expression of this gene in the SCN, where it was anticipated, Per expression was observed at ZT24 in the PT of the 129/SvEvBrd mice strain (23). Furthermore the expression in the PT was shown to be under circadian regulation as it continued under conditions of constant darkness (23). Although the secretion of melatonin in the 129/SvEvBrd strain of mice was not established, the authors showed that in C57 BL/6 mice, which have a genetic defect in melatonin production, there was no expression of Per1 in the PT (23). They concluded that melatonin present in the 129/SvEvBrd mouse strain was driving the rhythm of Per expression in the PT (23). If this conclusion is correct the data imply that melatonin may be driving Per expression in the PT during the hours of darkness, contrary to the prediction above. We have therefore investigated the expression of Per in sheep pars tuberalis as a marker of gene expression to define how nocturnal melatonin may be regulating its expression, and to test whether mela-tonin may be stimulating the PT gland.

A 404bp cDNA fragment of the ovine Per1 gene was generated by reverse-transcription PCR from mRNA isolated from sheep PT (21). This was subsequently cloned in pGEM-T, and sequenced to verify that a bone fide part of the sheep Per1 gene had been obtained. Alignment of the amplified region, 287-690 of the human Per1 gene (GenBANK accession number AB002107) showed strong homology to the human sequence (Figure 2). The cDNA probe was used to define the regulation and temporal expression of the oPer1 gene in sheep PT cells by Northern blot. In the absence of any stimulation, oPT cells displayed only a minimal signal, a single band at 4.9 kb, which did not change in expression over 48 h of culture. Forskolin, a diterpene used to stimulate increased cyclic AMP, increased oPer1 mRNA expression dramatically after 2-4h in culture, but this response returned to basal levels of expression within 6 h, indicating only a transient period of expression (Figure 3 and data not shown, see Morgan et al. 1998 (21)). Such transient expression is typical of early response genes, and consistent with this oPer1 was shown to increase dramatically in oPT cells a o D

Melatonin Receptor

Figure 2. Alignment of predicted protein fragment of oPerl. translated from ovine PCR product, with its corresponding region in human Perl. It shows the close similarity of this region at the amino acid level, with only 5 changes occuring over 134 amino acid. This confirmed that a fragment of sheep Perl had been amplified and cloned (GenBANK accession number AF044911).

Figure 2. Alignment of predicted protein fragment of oPerl. translated from ovine PCR product, with its corresponding region in human Perl. It shows the close similarity of this region at the amino acid level, with only 5 changes occuring over 134 amino acid. This confirmed that a fragment of sheep Perl had been amplified and cloned (GenBANK accession number AF044911).

RIGUI

G3DPH

G3DPH

Figure 3. Northern blot showing the expression of oPerl (4.9kb band) in ovine PT cells over a 48h period. G3PDH (1, 4kb band) provides a reference for loading. The data show how the expression of oPerl is increased after 4h of forskolin stimulation, but is relatively unaffected by any treatments at any other of the time points. (From Morgan et al. 1998 (21)).

in response to forskolin stimulation after 2 h even in the presence of the l0^g/ml cyclo-heximide. As cycloheximide is a protein synthesis inhibitor, the ability of oPerl mRNA expression to be increased in its presence demonstrates that de novo protein synthesis is not required for enhanced gene expression of the oPerl gene. These data confirm that oPerl is an early response gene (21).

A parallel Northern blot showed that the expression of oPerl in the pituitary is not restricted to the pars tuberalis. Expression within the pars distalis was also shown, although the response to stimulation by forskolin in the absence of cycloheximide was much weaker that in the PT. This probably reflects the weaker effect of forskolin of cyclic AMP levels in the PD that has been observed previously (12).

The phorbol ester, phorbol 12,13, myristate acetate, stimulates protein kinase C, and we have shown that it will stimulate c-fos expression in oPT cells (22) (Fig. 3). This powerful stimulus had no effect on the induction of oPerl expression, and therefore suggests that protein kinase C mediated pathways are not involved in the regulation of oPerl expression.

As we have shown previously for the induction of c-fos, another early response gene, melatonin had no effect of the induction of oPerl expression in oPT cells, but it strongly inhibited the forskolin-induced response, consistent with its known effect of preventing or reversing forskolin-induced cyclic AMP stimulation (data not shown). Therefore these data in ovine PT cell cultures are entirely consistent with all other data concerning the acute effects of melatonin, showing its ability to prevent cellular activation, yet having no independent effect over a 48 h period (see Fig. 3). Therefore these data do not support the notion that melatonin stimulates Perl expression response directly.

We next tested how oPerl expression changes in vivo. Six Suffolk cross-bred ewes were housed in a 12L : 12D photoperiod. Subsequently three were killed 3 h after lights off, and three 3 h after lights on. The expression of oPerl in both the PT and PD was measured by in situ hybridisation using an antisense 35S-riboprobe for oPerl followed by quantification by computing densitometry. Consistent with the Northern

Figure 4. Densitometric analysis of oPerl gene expression in sheep PT and PD, measured following in situ hybridisation using an 35S-antisense riboprobe to oPerl. (a) shows diurnal expression of oPerl in sheep pituitary measured 3 h after lights on (day) and 3 h after lights off (night).The sheep were maintained on a 12L: 12D photoperiod. (b) shows effect of photoperiodic background on oPerl expression in pituitary, where sheep were maintained on 16L:8D (LD) or 8L: 16D (SD). * indicates significant day-night or LD-SD difference (p < 0.001).

o tu

LD SD

LD SD

blot findings oPer1 expression was detected in both the PT and PD. However the level of expression was affected by the time of day they were killed. In both the PT and PD a significantly higher level of expression was observed in the sheep killed 3 h after lights on, relative to those sheep killed 3 h into the dark period (p < 0.001) (Figure 4). This indicates that there is a diurnal pattern of expression of oPer1 in both the PT and PD, and this infers a diurnal pattern of stimulation to both glands. To assess the effect of long and short days on the expression of oPer1 in the pituitary, Soay rams which had been entrained to the following photoperiodic cycles were used: 8 week long days (LD; 16L:8D) or 8 weeks short days (SD; 8L: 16D) (see Morgan et al. 1998 for details (21)). Animals from each photoperiod were killed 2 h after lights on, and then oPer1 expression assessed by in situ hybridisation and computing densitometry as above. A marked difference (p < 0.001) in the expression of oPer1 in the PT of LD and SD was observed, whereas no difference in expression was observed between the PD of LD and SD animals (Figure 4). This indicates that the difference in the level of oPer1 expression in the PT is due to the photoperiodic background, and hence mela-tonin. The lack of effect on the PD is explained, as there are no melatonin receptors on the PD.

These data allow several inferences to be drawn about how melatonin acts on the PT. Firstly, the increased level of oPer1 gene expression 3 h after lights on relative to 3h after lights off, in conjunction with the transient increase in expression of oPerl mRNA in PT cells within 2h following stimulation with forskolin, infers that a humoral stimulus drives the expression of oPerl in the pituitary (PT and PD) following lights on (and melatonin decline). From experiments using primary PT cell cultures we predict that when melatonin levels are high the PT will not be stimulated. Secondly the reduced amplitude in oPerl mRNA expression following a short relative to a long photoperiodic signal (long vs short duration melatonin) infers that the duration of the nocturnal melatonin signal influences either the amplitude of the response to the humoral stimulus or alters the timing of the peak. Recent work in Syrian hamsters measuring Perl expression at multiple time points shows that it is the amplitude of the peak in gene expression that is affected (unpublished observations). Thus these data provide the first evidence to show how a durational melatonin signal is interpreted by its target gland through gene expression. Namely duration is decoded into a signal of amplitude. Therefore while melatonin may be important in preventing cellular activation, these data suggest that an important aspect of melatonin signalling is a distinct "programming" effect related to its duration, and this influences subsequent cellular sensitivity to stimulation.

We have shown previously, using primary cell cultures, that melatonin alters the sensitivity of PT cells to produce cyclic AMP in response to stimulation by forskolin in a duration-dependent manner (4). This sensitisation process is pertussis-toxin sensitive (unpublished observations) and requires only physiological concentrations of mela-tonin (i.e. 100pM) (4), suggesting that the Mella receptor mediates this response. The mechanism involved is unknown, but it does not require protein synthesis, and is unlikely to involve any known signal transduction pathway (4). Nevertheless the results together with those above infer that melatonin, through its receptor, alters cellular sensitivity to stimulation.

There is however, a major discrepancy between the results produced in vivo and in vitro. In vivo the effect of short photoperiod (hence long duration melatonin) is to reduce the amplitude of oPerl expression relative to the long photoperiod (short duration melatonin) animals. In PT cells long duration melatonin signals (8-16 h) increase the amplitude of cyclic AMP production relative to short duration melatonin signals (0-4 h). Such an amplified response in cyclic AMP would be expected to amplify oPerl expression in short day animals, contrary to the abrogated response obtained. The reason for this apparent discrepancy is not immediately clear, and at the present time we cannot make all the parts of the jig-saw fit together. Nevertheless, both the in vivo and in vitro data strongly suggest that the mechanism through which the duration of the melatonin signal is relayed to the PT involves altered sensitivity of intracellular signalling pathways leading to a change in the amplitude of gene expression. Thus in terms of its biological function, the PT would be less biochemically active during short days than during long days.

The Insomnia Battle

The Insomnia Battle

Who Else Wants To Sleep From Lights Out 'Til Sunrise Without Staring At The Ceiling For Hours Leaving You Feeling Fresh And Ready To Face A New Day You know you should be asleep. You've dedicated the last three hours in the dark to trying to get some sleep. But you're wide awake.

Get My Free Ebook


Post a comment