| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratoire de Neurobiologie des Rythmes, Unité Mixte Recherche 7518, Centre National de la Recherche Scientifique/Université Louis Pasteur, Strasbourg, France
Abstract I. Introduction II. Role of Melatonin A. Regulation of Seasonal Rhythms B. Regulation of Circadian Rhythms C. Other Roles of Melatonin 1. Autocrine/Paracrine Effects. 2. Modulation of Neurotransmission 3. Effects of Melatonin on the Immune System. 4. Antioxidant/Antiaging Property of Melatonin. D. Sites and Mechanisms of Action of Melatonin E. Conclusion: Melatonin Is a Time-Giver Endocrine Messenger III. Neural and Humoral Inputs to the Mammalian Pineal Gland A. Structure and Ultrastructure of the Pineal Gland B. Neural Inputs 1. Retino-Hypothalamo-Pineal Pathway. a. The Retino-Hypothalamic Tract. b. The Hypothalamic Endogenous Circadian Oscillator. c. Suprachiasmatic Nucleus of the Hypothalamus Outputs to the Pineal Gland. 2. Central Pathways. 3. Parasympathetic Pathways. 4. Pathways from Other Neural Structures C. Endocrine Inputs D. Paracrine Inputs E. Conclusion: The Pineal Gland Is a Junction of Various Neural Inputs IV. Indoleamine Metabolism in the Mammalian Pineal Gland A. Indoleamine Metabolic Pathways B. Tryptophan Hydroxylase C. Aromatic Amino Acid Decarboxylase D. Monoamine Oxidase E. Alcohol and Aldehyde Dehydrogenases F. Arylalkylamine-N-Acetyltransferase G. Hydroxyindole-O-Methyltransferase V. Noradrenergic Regulation of Melatonin Synthesis in the Mammalian Pineal Gland A. Noradrenergic Regulation of Melatonin Synthesis in the Rat Pineal Gland 1. Adrenergic Receptors of the Pineal Gland a. Subtype {beta}1 b. Subtype {alpha}1 c. Subtype {alpha}2 2. Second Messengers Induced by Noradrenergic Stimulation. 3. The Third Messengers/Transcription Factors Induced by Noradrenergic Stimulation. 4. Acute Effects of Noradrenergic Stimulation on the Melatonin Synthesis Pathway. 5. Mechanisms Involved in the Termination of Nocturnal Melatonin Synthesis. 6. Effect of Light Exposure at Night. 7. Consequences of Long-Term Noradrenergic Stimulation of the Pineal Gland. B. Noradrenergic Regulation of Melatonin Synthesis in Other Mammalian Species 1. Daily Regulation of Melatonin Synthesis a. Daily Regulation of Melatonin Synthesis in Other Rodents. b. Daily Regulation of Melatonin Synthesis in Nonrodents. c. Conclusions. 2. Seasonal Variations in Melatonin Synthesis a. Variations in the Duration of the Nocturnal Melatonin Peak. b. Variations in the Amplitude of the Nocturnal Melatonin Peak. c. Conclusions. C. Conclusion: Both AA-NAT and HIOMT Shape the Daily and Seasonal Profiles in Melatonin Synthesis VI. Regulation of Melatonin synthesis in the Mammalian Pineal Gland by Other Transmitters A. Peptidergic Regulation of Melatonin Synthesis 1. Vasoactive Intestinal Peptide, Pituitary Adenylate Cyclase Activating Peptide, and Histidine Isoleucine Peptide. 2. Neuropeptide Y. 3. Vasopressin and Oxytocin. 4. Somatostatin. 5. Substance P. 6. Calcitonin Gene-Related Peptide. 7. Secretoneurin. 8. Hypocretin. 9. Delta-Sleep Inducing Peptide. 10. Natriuretic Peptides. 11. Angiotensin. 12. Opiate Peptides. 13. Luteinizing Hormone-Releasing Hormone. 14. Peptides to Come. 15. Conclusion: (Neuro)Peptides Are True Pineal Transmitters. B. Other Nonadrenergic, Nonpeptidergic Transmitters of the Pineal Gland 1. Serotonin. 2. Dopamine. 3. Acetylcholine. 4. Glutamate. 5. GABA. 6. Taurine. 7. Histamine. 8. Adenosine and ATP. 9. Nitric Oxide. 10. Gonadal Steroids. VII. General Conclusions and Perspectives
 |
Abstract |
|---|
 Top Next |
|---|
|
I. Introduction |
|---|
|
TopPreviousNext |
|---|
The pineal gland was studied intensively by Descartes during the 17th century. He described the pineal gland as the third eye, not by analogy to its role in the control of the photoperiod, which he had no knowledge of, but because it is, in the Cartesian dualist vision, the place in the body where the soul exerts its control (the seat of imagination and common sense), and not the seat of the soul as it has often been referred to. "The reasonable soul," according to Descartes, "is lodged in the body, but not only as a pilot on its ship, it is necessary that it is united with its body." Descartes was the first to propose a "physiological" explanation for the functioning of the central nervous system, including the pineal gland, for the perception of the environment. Even if this Cartesian model appears a posteriori an unreliable model, this concept nevertheless prevailed for the next 250 years.
At the end of the 19th century Ahlborn and Rabl-Ruckhardt, then Graaf, Korschelt, and Spencer, described the anatomy, histology, innervation, and embryology of the mammalian pineal gland and noticed its resemblance to the epiphysis organ of lower vertebrates. In 1905, Studnicka established that phylogenetically the pineal gland derived from a photoreceptor organ, but its function remained unknown.
At the beginning of the 20th century the physiological role of the pineal gland was studied. Heubner presented the case of three girls with pineal tumors and precocious puberty. He concluded that the destruction of the pineal by the tumor had prevented the normal production of an antigonadotropic pineal hormone and raised the hypothesis that the pineal may control the onset of puberty. The link between the pineal gland and reproduction was thus established. In 1943, Bargman suggested that the endocrine function of the pineal gland was regulated by light, via the central nervous system.
From the 1970s, the number of publications on the pineal gland markedly increased. The first international congress that brought "pinealogists" together was held in 1965 in Amsterdam. Research on the pineal gland developed in four main directions.
|
The objective of this review is to consolidate and update our current knowledge of the complex and varied inputs controlling the rhythmic synthesis of MEL in the mammalian pineal gland.
|
II. Role of Melatonin |
|---|
|
TopPreviousNext |
|---|
; Goldman, 1999 for reviews). For many years, but especially during the last decade, many studies have been performed to understand the physiological role, sites, and mechanisms of action of MEL. A. Regulation of Seasonal Rhythms
The pineal gland is a major component of the endocrine system that allows mammals to respond to the annual changes in photoperiod by adaptive alterations of their physiological state. The best example of such photoperiod-dependent physiological functions is the activation/inactivation of the reproductive axis, a phenomenon in which the pineal and its MEL rhythm are essential. Numerous studies have now demonstrated that the pineal gland is a neuroendocrine transducer receiving photoperiodic information from the retina and circadian SCN oscillator, and transmitting this to the reproductive system via a particular dynamic pattern of MEL secretion (see Hoffmann, 1979; Reiter, 1980; Goldman and Darrow, 1983; Bittman, 1984; Tamarkin et al., 1985; Pévet, 1988; Goldman, 2001 for reviews). However, several fundamental questions remain before the role of MEL in the regulation of seasonal function is elucidated: 1) where is the photoperiodic information encoded before its translation into the MEL rhythm? 2) Where and how is the MEL rhythm decoded to regulate specific seasonal functions? 3) Which parameters of the MEL rhythm (phase, duration, amplitude, or total quantity) are interpreted as the photoperiodic message by the target structures?
Recently, data have accumulated that strongly suggest that the hypothalamic circadian clock may be the site for the integration of annual changes in photoperiod (see Goldman, 2001; Schwartz et al., 2001 for reviews): namely, a circadian reading of the photoperiod appears necessary (Maywood et al., 1990); FOS reactivity in the SCN following a light stimulus depends on the photoperiod history (Sumova et al., 1995; Vuillez et al., 1996); clock gene expression in the SCN displays MEL-independent photoperiodic variations (Messager et al., 1999b, 2000, 2001; Nuesslein-Hildesheim et al., 2000); and the daily profile of vasopressin (VP) mRNA differs in long and short photoperiods (Jac et al., 2000). In addition, the thalamic intergeniculate leaflet (IGL), a relay between the retina and SCN, may be involved in photoperiod integration (Menet et al., 2001).
Several neural structures have been identified as targets for MEL's effect on seasonal function. The pars tuberalis of the adenohypophysis (PT), containing the highest density in MEL receptors (MEL-R), is the site of action for MEL regulation of prolactin secretion (see Lincoln, 1994; Malpaux et al., 1995, 2001; Hazlerigg et al., 2001 for reviews) and displays MEL-dependent daily and photoperiodic variations in clock gene expression with lower amplitude under a short photoperiod (SP) (Messager et al., 1999b, 2000, 2001; von Gall et al., 2002a). Identification of the specific molecule released from the PT in response to MEL, which acts on the lactotrophs, named tuberalin, remains unknown, although two 21- and 72-kDa proteins were recently identified in the bovine PT (Guerra and Rodriguez, 2001). Depending on the species, various hypothalamic sites (SCN in Siberian hamster; mediobasal hypothalamus in Syrian hamster, premammillary hypothalamus in sheep) are MEL targets for the specific control of reproductive function (Badura and Goldman, 1992; Maywood and Hastings, 1995; Malpaux et al., 1998). Although it has been clearly shown that MEL is the photoperiodic endocrine message for each structure, it has not yet been elucidated how this MEL message is decoded at the cellular level. Several studies have reported that, although MEL is an acute inhibitor of cAMP accumulation, tissues pre-exposed to long-duration (up to 16 h) MEL treatment become hypersensitive to cAMP (Hazlerigg et al., 1993; Witt-Enderby et al., 1998; Messager et al., 1999a; Pelisek and Vanecek, 2000) or cAMP elevating agents like adenosine (von Gall et al., 2002a) even with a lower number of MEL-R.
To define which parameters of the MEL secretion pattern (phase, duration, amplitude, or total quantity) are interpreted as a photoperiodic message by the target structures, several hypotheses have been proposed from analysis of the endogenous MEL patterns in different conditions and from studies with acute injections or chronic infusions of exogenous MEL (Fig. 2). Observations of the MEL secretion pattern in various species raised in different photoperiodic conditions have shown that the duration of the nocturnal MEL peak is positively related to the length of the night (sheep: Rollag and Niswender, 1976; Karsch et al., 1988; rat: Illnerova and Vanecek, 1980; Siberian hamster: Illnerova et al., 1984; Ribelayga et al., 2000; Syrian hamster: Skene et al., 1987; Maywood et al., 1993; Miguez et al., 1995a; European hamster: Vivien-Roels et al., 1992). Furthermore, experiments using acute injections or constant infusion of MEL have shown that the duration of a high circulating MEL level is the limiting factor to obtain a photoperiodic response (see Carter and Goldman, 1983; Pitrosky et al., 1991; Bartness et al., 1993 for reviews). Consequently, the duration of the nocturnal MEL peak is an important factor for the transmission of photoperiodic information from the environment to the body. The early experiments showed that an acute injection of MEL at the end of the day or beginning of the night to intact hamsters kept in long photoperiod (LP) induced gonadal regression, while a similar injection made at the end of the night or at the beginning of the day had no effect. This observation led to the hypothesis that the coincidence of the injection of MEL with a phase of sensitivity was a deciding factor for the appearance of a physiological effect (see Tamarkin et al., 1976; Reiter, 1987 for review). Recently, a study performed in our laboratory (Pitrosky et al., 1995) has shown that the photoperiodic response to MEL in the Syrian hamster depends on a phenomenon of coincidence. The infusion of two consecutive MEL peaks, whose length from the beginning of the first peak to the end of the second peak corresponded to an SP signal but whose total quantity of infused MEL corresponded to an LP signal, induced an SP-type response of the reproductive axis. The physiological response thus depends on the interval between the first and the second MEL peak but not at the clock time when the double MEL peak is applied.
|
In addition, the amplitude of the nocturnal peak of MEL could also be an important parameter in photoperiodic transmission (see Vivien-Roels, 1999 for review). Several examples of photoperiodic variation in the amplitude of the MEL peak have been observed, for example, in the pig (McConnell and Ellendorf, 1987; Taste et al., 2001), mule (Cozzi et al., 1991), Siberian hamster (Lerchl and Schlatt, 1992; Steinlechner et al., 1995; Miguez et al., 1996; Ribelayga et al., 2000), European hamster (Vivien-Roels et al., 1992, 1997), and horse (Guérin et al., 1995). Annual variations in the amplitude of the nocturnal MEL peak are especially visible when animals are maintained in their natural environment. These observations suggest that factors other than the photoperiod that display annual variations (e.g., temperature, quality/quantity of food, humidity) may be integrated by the organism and transmitted via the secretion of MEL (Pévet, 1987; Pévet et al., 1991; Vivien-Roels, 1999). These other nonphotic environmental factors could modulate the perception of the photoperiod by altering the metabolism of the pineal gland. Environmental temperature seems an important factor since diminution of the temperature accelerates gonadal regression in Siberian and Syrian hamsters placed in SP (Heldmaier and Steinlechner, 1981; Pévet et al., 1986; Larkin et al., 2002). In addition, a decrease in temperature 1) increases enzyme activity in the rat pineal gland (Nir et al., 1975); 2) increases the amplitude of the nocturnal pineal MEL peak in the Syrian hamster (Brainard et al., 1982, but discussed by Pévet et al., 1989a) and European hamster (Vivien-Roels et al., 1997); and 3) modulates the inhibitory effect of light applied at night (Stieglitz et al., 1991). Currently, anatomical structures and transmitters involved in these effects of temperature are not known and could act directly on the pineal gland or on intermediate structures sensitive to the temperature.
Historically considered as a pro or antigonadotropic hormone, according to species, it is clearly established now that MEL is a pivotal endocrine messenger used to time several annual functions with the seasonal cycle to ensure adaptation and survival of individual in their cyclic environment.
B. Regulation of Circadian Rhythms
In all mammals studied to date, whether they exhibit nocturnal or diurnal activity, MEL is synthesized in the pineal gland during the dark phase of the light/dark cycle and is rapidly delivered to the body via the bloodstream. Pinealectomy does not alter the animal's circadian rhythm in rest-activity but facilitates the re-synchronization of the animal to a new photoperiod (Cheung and McCormack, 1982). The daily rhythm of MEL is considered to be a circadian mediator used by the endogenous SCN clock to deliver the circadian message to MEL target structures (containing MEL-R). In addition, MEL exerts a "chronobiotic" effect by acting directly on the SCN, which contain MEL-R (Vanecek et al., 1987), to affect the circadian clock (see Pévet et al., 2002 for review).
In rats and hamsters with free-running circadian rhythms, pharmacological doses of exogenous MEL are capable of synchronizing the circadian rhythms of locomotor activity and MEL synthesis (see Redman et al., 1983; Armstrong and Chessworth, 1987; Humlova and Illnerova, 1990; Kirsch et al., 1993; Drijfhout et al., 1996b; Grosse and Davis, 1998; Pitrosky et al., 1999; Schuhler et al., 2002; Pévet et al., 2002 for review). The synchronizing effect of MEL occurs at a particular circadian time, being different according to species (e.g., beginning of the active period, CT 12, in the rat). Recently, it was reported that exogenous MEL, applied directly into the SCN by reverse microdialysis, not only phase-advances the endogenous MEL peak but also increases the amplitude of the MEL peak (Bothorel et al., 2002). Additionally, various in vitro studies have demonstrated a local effect of MEL on SCN metabolism, electrical activity, and circadian rhythmicity (Cassone et al., 1988; Stehle et al., 1989; Mc Arthur et al., 1991). At the moment, it is not known why high doses of exogenous MEL are necessary to induce a phase-shifting effect. MEL may exert its synchronizing properties indirectly on clock inputs or clock outputs, or directly on the clock via MEL-R (MEL-R were identified on VP-containing SCN neurons; Song et al., 1999) or other binding sites (see Pévet et al., 2002 for review). This property of MEL is used, along with several circadian signals, between the mother and fetus to entrain the circadian clock of the offspring (Reppert et al., 1979; Reppert and Weaver, 1991).
In humans, this "chronobiotic" property of MEL has been used to help re-synchronize individuals showing disrupted circadian rhythms, for example, related to "delayed sleep phase" syndrome, jet-lag, night shift work, or in some blind people (Arendt et al., 1984, 1987, 1988, 1997; Lewy et al., 1992; Claustrat et al., 1995; Skene et al., 1996; Lockley et al., 2000; Takahashi et al., 2000).
1. Autocrine/Paracrine Effects.
In addition to the pineal gland, MEL is synthesized in several other structures (retina, Harderian gland, gut) where the genetic expression and biochemical activity of the MEL-synthesizing enzymes have been detected (Quay, 1965; Cardinali and Wurtman, 1972; Quay and Ma, 1976; Brammer et al., 1978; Pévet et al., 1980a; Vivien-Roels et al., 1981; Gauer and Craft, 1996; Roseboom et al., 1996; Ribelayga et al., 1998a; Djéridane et al., 1998, 2000). Since following pinealectomy the plasma MEL concentration is very low and since some of these structures contain MEL-R (Dubocovich and Takahashi, 1987; Lopez-Gonzalez et al., 1991), it has been proposed that MEL plays an auto/paracrine role in these structures.
In the retina, MEL is rhythmically synthesized in the photoreceptors in a circadian manner (see Cahill and Besharse, 1995 for review), which persists in vitro in constant conditions (Tosini and Menaker, 1996, 1998). MEL alters various aspects of retinal metabolism (see Iuvone, 1996 for review). Most of the retinal effects of MEL are indirect, and probably consist primarily in the inhibition of dopamine (DA) release from amacrine cells (Dubocovich, 1983). Conversely, DA acutely inhibits MEL synthesis in the retina and affects the phase of the MEL rhythm (Iuvone et al., 1987; Nguyen-Legros et al., 1996; Jaliffa et al., 2000; Tosini and Dirden, 2000).
The rodent Harderian gland also synthesizes MEL but the mechanisms regulating the synthesis and local effects of the hormone are still not well understood (Djéridane et al., 1998, 2000).
In the pineal gland several observations also suggest that MEL exhibits autocrine/paracrine effects. For example, in neonate but not adult rats, the pineal gland displays MEL-R binding (Zitouni et al., 1995). Exogenous MEL modifies various morphological and biochemical pineal parameters, namely proteic microtubules (Freire and Cardinali, 1975), enzymatic activities (Freire and Cardinali, 1975), presynaptic release of neurotransmitters (Chuluyan et al., 1991), and pre and postsynaptic release of the MEL precursor serotonin (5-HT; Miguez et al., 1995b).
2. Modulation of Neurotransmission
It has been proposed that MEL could, on one hand, alter the release of several neurotransmitters, especially DA, 5-HT, norepinephrine (NE), acetylcholine (ACh) and, on the other hand, could modulate the postsynaptic response (Cardinali et al., 1975; Carneiro et al., 1994; Markus et al., 1996; Bucher et al., 1999). For example, MEL potentiates the NE-induced vasoconstriction of the rat caudal artery (Bucher et al., 1999). In addition, MEL, through activation of its different receptor subtypes, can differentially modulate the function of type A 
-aminobutyric acid (GABAA) receptors (Wan et al., 1999). It has been proposed that some effects of exogenous MEL in humans (sedative, analgesic, anticonvulsive, anxiolytic) could be related to its interaction with the GABAergic system (Golombek et al., 1996).
3. Effects of Melatonin on the Immune System.
Earlier studies reporting that pinealectomized rats displayed a structurally modified thymus and that MEL treatment or pineal grafting prevented thymic involution in very old mice led to the concept that MEL could affect the immune system (see Provinciali et al., 1996; Liebmann et al., 1997; Reiter et al., 2000a; Maestroni, 2001 for reviews). In vivo, high exogenous doses of MEL show a general stimulation of the immune system. It increases T cell activity, lymphocyte growth, humoral responses, and may inhibit thymus involution with age. In vitro MEL also increases T helper and NK cell activities, the production of interleukin 2 and interferon gamma, and the expression of interleukin 1 mRNA in human monocytes. In summary, most authors agree on an immuno-stimulating effect of MEL. These effects may occur via a direct action of MEL on its receptor since MEL-R have been identified in various tissues of the immune system, namely thymus, spleen, lymphocytes, and T helper cells.
In addition, MEL acting as a chronobiotic may be involved in the circadian organization of the immune system (the number and activity of lymphocytes T, B, and NK displaying a daily rhythmicity). It has also been proposed that MEL may mediate seasonal changes in immune function, which is enhanced in short days with longer MEL peak duration (Nelson and Drazen, 2000).
4. Antioxidant/Antiaging Property of Melatonin.
The publication of a revitalizing effect of MEL or of pineal youth transplants to old mice (Pierpaoli and Regelson, 1994) raised a general interest for MEL as an antiaging/antioxidant molecule. It was proposed that the lipophilic MEL diffuses into the cell cytosol and nucleus (Menendez-Pelaez and Reiter, 1993) to protect cytosolic and nuclear macromolecules from free radical cytotoxicity (see Reiter, 1995; Reiter et al., 2000b for reviews).
The use of oxygen in cell metabolism leads to the production of cytotoxic by-products that are reactive free radical species (superoxide anion radical, peroxynitrite anion, hydrogen peroxide, nitric oxide, and the highly toxic hydroxyl radical), which destroy macromolecules like DNA, lipids, and proteins leading to cell death via apoptosis. High doses of MEL (in the micromolar range) are reported to neutralize most of these cytotoxic molecules, but especially the hydroxyl radical, which is scavenged in vivo by MEL, producing cyclic 3-hydroxyMEL excreted in the urine. In addition, MEL is reported to stimulate the activity of various antioxidant enzymes, like superoxide dismutase or glutathione peroxidase, but inhibits the pro-oxidant enzyme nitric oxide synthetase.
Given that MEL could be a very powerful antioxidant molecule, that the production of MEL decreases with age (although this conception is now discussed, see Kennaway et al., 1999), and that the free radical effects are involved in the processes of aging and cancer, it has been suggested that maintaining MEL at a high level could slow age- and cancer-related alterations (Reiter, 1995; Reiter et al., 2000b). The anticarcinogenic effect of MEL is best described in vivo and in vitro on the estrogen-responsive mammary tumors (Tamarkin et al., 1981; Blask and Hill, 1986; Hill and Blask, 1988; Scott et al., 2001; Teplitzky et al., 2001; Kiefer et al., 2002). In vivo, there is an inverse correlation between the nocturnal level of plasma MEL and the number of estrogen receptors in patients with an estrogen-dependent cancer. In vitro, 1 to 100 nM MEL induces a 40 to 60% loss of MCF-7 cells (human breast tumoral cells). This cytotoxic effect of MEL is related to an apparent uncoupling of oxidative phosphorylation and leads to morphologic alteration and autophagocytosis. MEL also affects estrogen receptor transcriptional activity by regulating signal transduction pathways. In addition, MEL has been described as a potent supplement in the treatment or cotreatment of cancer: as an antioxidant, it may protect cell damage caused by carcinogens; as a chronobiotic, it may help determine optimum timing for carcinogen best efficiency; and it may act in synergy with the carcinogen retinoic acid to markedly reduce mammary tumor genesis in vivo.
It is noteworthy that most of these effects necessitate pharmacological doses of MEL (in the micromolar range) while plasma MEL concentrations are in the picomolar range. Recent studies, however, suggest that MEL could display antioxidant properties even at physiological levels (Pozo et al., 1994; Benot et al., 1999). Nevertheless, even if used at high doses, the therapeutic effect of MEL should not be neglected. Additionally, it is proposed that MEL could also serve to maintain synchronization of the main biological functions and prevent disintegration of the circadian oscillator in the course of aging (Armstrong and Redman, 1991).
D. Sites and Mechanisms of Action of Melatonin
The hormonal MEL message delivered by the pineal gland is distributed rapidly via the systemic circulation to all peripheral and central structures where MEL acts via specific binding sites (see Weaver et al., 1991; Masson-Pévet et al., 1994a, 1996; Morgan et al., 1994; Vanecek, 1998; von Gall et al., 2002b for reviews).
The localization and pharmacological characterization of the MEL binding sites were made possible in 1987 with the use of a radioiodinated MEL ligand (125I-MEL, Vanecek et al., 1987). Two types of binding sites have been characterized: the high-affinity sites (with a constant of dissociation (KD) between 20 and 200 pM), and the low-affinity sites (with a KD value in the nanomolar range). Only the high-affinity sites have been characterized as receptors (MEL-R), and their genes cloned. Three types of high-affinity receptors have been characterized (see Reppert et al., 1996, for review; Dubocovich et al., 2001 for latest nomenclature): MT1 (previously Mel1a) present in all vertebrates, mainly in the brain; MT2 (previously Mel1b) present in all vertebrates, mainly in the retina; and Mel1c, present in nonmammalian vertebrates. The low-affinity binding sites, MT3, were recently described as the quinone reductase 2 enzyme (Nosjean et al., 2000).
The MT1 receptor has seven transmembrane domains, specific to G-protein-coupled receptors, and are coupled negatively to the adenylate cyclase (AC) system. Their activation induces a decrease in forskolin-induced cAMP accumulation (Carlson et al., 1989; Morgan et al., 1989). This effect is generally mediated by a pertussis toxin-sensitive G-protein (Gi/Go; Reppert et al., 1994). In the PT, MEL-Rs are coupled to two types of G-proteins, one sensitive to the pertussis toxin, the other to the cholera toxin. Other effects of MT1 activation have also been reported on the intracellular concentrations of cGMP, diacylglycerol (DAG), inositol triphosphate (IP3), or Ca2+; on the activity of protein kinase Ca2+ and/or DAG-dependent (PKC); on the expression of c-fos; on the phosphorylation of cAMP responsive element (CRE)-binding protein (CREB); and on membrane potential.
Currently, about 110 cerebral structures express MEL binding sites. The number and nature of these structures display marked interspecific variations. In nearly all mammals, the SCN mainly express MT1 receptors with the exception of the mink and sheep. The PT is an endocrine structure characterized by a very high density of MT1 receptors in all mammals except humans. The MT2 receptor is present in the retina (Reppert et al., 1995) and possibly in the brain and SCN as well (Dubocovich et al., 1998; Isobe et al., 2001). In the SCN, MT1 receptors would mediate the inhibitory effect on electrical activity, whereas the MT2 receptor would mediate the phase-shifting effect of MEL. Notably, a nonsense mutation occurs in the MT2 coding gene in Siberian and Syrian hamsters, which disables the receptor (Weaver et al., 1996). MEL-Rs are present in the pineal gland of the newborn rat, become rare in 9-day-old rats, and are not detected in adults (Zitouni et al., 1995). MEL-Rs have also been characterized in many peripheral structures such as the Harderian gland, spleen, testis, ovary, vascular system, gut, smooth muscle, and some cells of the immune system (see Vanecek, 1998 for review).
E. Conclusion: Melatonin Is a Time-Giver Endocrine Messenger
MEL is a time-giver (zeitgeber) hormone. It is characterized by two rhythms of secretion: a 24-h rhythm with a nocturnal peak and an annual rhythm closely dependent on seasonal variations in the photoperiod. It is possible that most, if not all, functions attributed to MEL are related to the timing information it brings to different structures. Studies performed to understand the mechanisms of action of MEL in the regulation of some seasonal and circadian functions have demonstrated that the dynamic pattern of MEL secretion is fundamental for its time-giving function. The rhythmic pattern of MEL secretion is important because it brings to organisms information about time that allows them to adapt some of their physiological functions to the daily and seasonal variations of their environment. It is thus necessary to delineate the various processes and elements that regulate the rhythms of MEL synthesis and secretion to understand how environmental factors are transmitted to the whole organism.
|
III. Neural and Humoral Inputs to the Mammalian Pineal Gland |
|---|
|
TopPreviousNext |
|---|
A. Structure and Ultrastructure of the Pineal Gland
The mammalian pineal structure and ultrastructure have been largely described in previous reviews (Vollrath, 1981; Pévet, 1983a). The pineal gland develops as an evagination of the diencephalic roof. In most mammals it forms a solid mass between the habenular and posterior commissures, but in rodents, whereas a deep and small part stays close to ventricle III, the major portion of the gland migrates in a dorso-caudal direction to form the superficial pineal, both parts being connected by the pineal stalk (see Fig. 1). The rodent superficial pineal gland is massively innervated and contains a dense network of blood vessels into which MEL is released. However, in the deep pineal gland, being made of functional pinealocytes that express the genes coding for the MEL-synthesizing enzyme with a day/night rhythm (Ribelayga et al., 1998a; Garidou et al., 2001), MEL could as well be directly released into the cerebrospinal fluid, as has been recently demonstrated in sheep (Tricoire et al., 2002). In the course of phylogenesis, the pineal gland has undergone marked transformations (Collin, 1971; Korf et al., 1998). Being made of true photoreceptors in lower vertebrates, in mammals it consists of neuroendocrine cells, the pinealocytes, with no direct light sensitivity but still expressing various photoreceptor markers (rhodopsin, S-antigen, recoverin, etc.). The mammalian pineal gland is a rather homogenous tissue containing mainly true pinealocytes (mono-, bi-, or tri-polar cells), few glial cells, phagocytic cells, and rare neurons.
The pineal gland is innervated with nervous fibers of various origins (Fig. 3). The main pathway consists of a complex route named the retino-hypothalamo-pineal pathway, ending with the sympathetic input to the pineal parenchyma. The pineal gland also receives neural inputs of central and parasympathetic origins. These pineal nerve endings contain a large variety of neurotransmitters.
|
1. Retino-Hypothalamo-Pineal Pathway.
The rhythm in MEL synthesis depends essentially upon three interdependent factors: the endogenous circadian oscillator located in the SCN, the L/D cycle that synchronizes the endogenous oscillator, and the light that acutely inhibits nocturnal MEL synthesis. It is now well established that there exists a multi-synaptic neural pathway among the retina, SCN, and pineal gland. Various experiments (lesion, neuronal tracing) have allowed researchers to draw the general diagram of the main innervation of the mammalian pineal gland, especially in the rat (Moore and Klein, 1974; Klein and Moore, 1979; Moore, 1996; Larsen, 1999; Teclemariam-Mesbah et al., 1999).
a. The Retino-Hypothalamic Tract.
Photic information is conveyed from the retina to the ventro-lateral zone of the SCN via the retino-hypothalamic tract (RHT). The light-sensitive cells forwarding the light/dark information do not appear to be the rod and cone photoreceptors (Lucas et al., 1999), but rather are a small subset of retinal ganglion cells containing the photopigment melanopsin (Moore et al., 1995; Berson et al., 2002; Hannibal et al., 2002; Hattar et al., 2002). The RHT neurotransmitters are mainly glutamate (Glu) (van den Pol, 1991; Ding et al., 1997) and pituitary adenylate cyclase activating peptide (PACAP) (Hannibal et al., 1997), but not substance P (sP), as previously thought (Takatsuji et al., 1991) mediating light signaling to the clock (see Hannibal, 2002 for review). Other inputs originating from the thalamic IGL, containing neuropeptide Y (NPY), enkephalin (Enk), and GABA (Card and Moore, 1982; Moore and Speh, 1993; Morin and Blanchard, 2001) and from the raphe nucleus, containing 5-HT (Moore et al., 1978) also carry photic and nonphotic information to the SCN (Mrosovsky, 1996).
b. The Hypothalamic Endogenous Circadian Oscillator.
In mammals, several experiments have demonstrated the presence of an endogenous circadian oscillator in the SCN (see Ralph et al., 1990; Takahashi, 1995; Turek et al., 1995 for reviews) probably located in every SCN neuron showing an endogenous oscillation in firing rate (Welsh et al., 1995). This endogenous activity is higher during the subjective day and synchronized to exactly 24 h by the photic inputs. The cellular and molecular basis of this circadian oscillation and its synchronization are currently under active investigation. Several proteins (PER1–3, TIM, CLOCK, BMAL/MOP3, TAU/type I
casein kinase, cryptochrome 1–2) work as transcription factors and enzymatic regulators in an autoregulatory transcriptional feedback loop constituting the core of the circadian pacemaker (see Whitmore et al., 1998, 2000; Dunlap, 1999; Ishida et al., 1999; Kume et al., 1999; King and Takahashi, 2000; Lowrey and Takahashi, 2000; van Esseveldt et al., 2000; Reppert and Weaver, 2001 for reviews). Other elements of the circadian clockwork are still being discovered. The central step in transducing the intracellular cycling of molecular clocks to the rhythm in spontaneous firing rate was recently demonstrated to involve L-type Ca2+ current (Pennartz et al., 2002). SCN neurons are mainly peptidergic cells containing vasoactive intestinal peptide (VIP), VP, gastrin-releasing peptide (GRP), and somatostatin (SOM), but also GABA (see Buijs et al., 1994, 1995; Inouye, 1996; van Esseveldt et al., 2000 for reviews). Some of the peptides in the SCN display daily and/or circadian rhythms in their synthesis and release, thus being putative clock outputs.
It is suggested that the hypothalamic clock could also be involved in the integration of seasonal information (see Pittendrigh and Daan, 1976; Illnerova and Vanecek, 1985, 1987; Pévet et al., 1996; Goldman, 2001; Hastings, 2001; Schwartz et al., 2001 for reviews). For example, FOS-light induction (Sumova et al., 1995; Vuillez et al., 1996; Jacob et al., 1997) and Per1 gene expression (Messager et al., 1999b, 2000, 2001; Nuesslein-Hildesheim et al., 2000) in the SCN displays MEL-independent photoperiodic variations. In addition, the daily profile of vp- mRNA differs in long and short photoperiods (Jac et al., 2000). The integration of the photoperiod by the SCN has been proposed to involve two components (one recognizing variations of the dawn, the other of the dusk) with the increase (in the evening) and the diminution (in the morning) of MEL synthesis being regulated separately during photoperiod changes. The phase relationship between these two oscillator components would determine the duration of the nocturnal MEL peak (Illnerova and Vanecek, 1985, 1987). Recent observations in cultured SCN slices of Syrian hamsters have brought anatomical evidence for this concept (Jagota et al., 2000). However, an alternative view proposes that the photoperiod may be integrated into every SCN cell, into the molecular mechanism of the circadian clock itself. By affecting the daily profile of the light-sensitive Per expression (long under LP, short under SP), photoperiod may, in turn, affect the kinetics of the expression of the clock proteins and consequently the expression of all the clock-regulated genes (see Hastings, 2001 for review). Although it has been demonstrated that photoperiod clearly regulates the daily profile of Per1 (Messager et al., 2000) and PER1 (Nuesslein-Hildesheim et al., 2000) in the SCN, the link between changes in the clock-gene expression profile and SCN outputs remains to be established.
c. Suprachiasmatic Nucleus of the Hypothalamus Outputs to the Pineal Gland.
Many studies seek to elucidate how the temporal information generated by the SCN is transmitted to the organism to regulate many rhythmic physiological and behavioral functions (see Buijs, 1996; Buijs and Kalsbeek, 2001; Kalsbeek and Buijs, 2002 for reviews). It is generally considered that the ventro-lateral part of the SCN is the clock input area for the synchronizing events while the dorso-median part contains the oscillator and the output of the timing information. Actually, various SCN neurons project mainly to different hypothalamic structures to transmit the timing information to different functional axes, especially the hypothalamo-pituitary-adrenal axis (rhythmic secretion of corticosterone) and the hypothalamopineal axis (rhythmic secretion of MEL). Recently, the link between the SCN output and the circadian rhythm in locomotor activity was proposed to be the transforming growth factor 
acting on the hypothalamic subpara-ventricular zone (Kramer et al., 2001). In addition, the SCN could regulate peripheral endocrine organs via the autonomic nervous system (Buijs et al., 1999, 2001; Kalsbeek et al., 2000a; La Fleur et al., 2000). The increasing use of cDNA microarrays will help to identify new clock-controlled genes in various tissues (Akhtar et al., 2002; Duffield et al., 2002; Humphries et al., 2002).
In the rat, the SCN neurotransmitters involved in the clock output would be essentially VP and GABA (Moore and Speh, 1993; Buijs et al., 1994; Kalsbeek et al., 1995; 1996a). VP appears to be a good clock-controlled transmitter since 1) it displays a circadian rhythm of synthesis and release (Reppert, 1985; Murakami et al., 1991; Kalsbeek et al., 1995; Watanabe et al., 2000); 2) its gene promoter, containing an "E-box," is under the direct control of the clock genes (Jin et al., 1999); and 3) it acts on the dorsomedial hypothalamus to control the circadian rhythm of corticosterone synthesis and release (Kalsbeek et al., 1996b). In addition, VIP (Teclemariam-Mesbah et al., 1997a), glutamate (Cui et al., 2001), or another unknown diffusible substance (Silver et al., 1996; Allen et al., 2001) may also be non-neural outputs of the molecular clock.
As far as the regulation of MEL synthesis is concerned, the hypothalamic paraventricular nuclei (PVN) are an essential relay between the SCN and the pineal gland. PVN lesions abolish the rhythm of MEL synthesis in the pineal gland (Klein et al., 1983), PVN neurons respond to an electrical stimulation of SCN cells (Hermes et al., 1997), VIP or VP infusion in the PVN elevates pineal melatonin release (Kalsbeek et al., 1993), and retrograde labeling from the pineal gland is seen in the PVN (Larsen, 1999; Teclemariam-Mesbah et al., 1999). GABA appears to be involved in transmitting signals from the SCN to the PVN since infusion of a GABA antagonist during the subjective day in the PVN area stimulates MEL synthesis, whereas infusion of GABA during the night inhibits nighttime MEL secretion (Kalsbeek et al., 1996a, 1999, 2000b). SCN lesions abolish the daily rhythm of MEL synthesis but keep MEL at a level intermediate between daytime and nighttime values. These data indicate that the SCN is a daytime inhibitor (via GABA) of the PVN stimulation of MEL synthesis, and is probably also a nighttime stimulator (Kalsbeek et al., 2000b).
The dorsal and lateral parvocellular neurons of the PVN, containing oxytocin (OT) and VP, reach the intermediolateral cells (IML) of the upper three segments of the spinal cord (Gilbey et al., 1982; Yamashita et al., 1984; Cechetto and Sapper, 1988; Teclemariam-Mesbah et al., 1997b; Larsen, 1999). Diurnal inhibition of pineal gland activity could also take place at this level since 1) infusion of VP and especially OT in the IML inhibits the electrical activity of the preganglionic neurons of the spinal cord (Gilbey et al., 1982); and 2) inhibition of MEL synthesis following PVN electrical stimulation (Reuss et al., 1985; Olcese et al., 1987) is abolished in VP-deficient Brattleboro rats (Reuss et al., 1990). The IML neurons innervate the rostral pole of the superior cervical ganglion (SCG) neurons that project to the pineal gland (Strack et al., 1988; Reuss et al., 1989). This last step is excitatory since electrical SCG stimulation increases MEL release (Bowers and Zigmond, 1980). ACh is the main neurotransmitter released in the SCG (Kasa et al., 1991), but other neurotransmitters, especially SOM, VIP, histidine isoleucine peptide (PHI), and calcitonin gene-related peptide (CGRP) are potential candidates in the transmission of information to the SCG. Approximately 0.5 to 1% of SCG neurons project to the pineal gland (Bowers et al., 1984; Larsen, 1999).
The mammalian pineal gland is characterized by a very dense sympathetic innervation (see Kappers, 1960; Korf, 1996; Møller, 1999; for reviews). The first demonstration of the presence of neurotransmitters in the rat pineal gland was made using the technique of Falck et al. (1962), which showed the presence of NE in the sympathetic fibers of the pineal gland. In the rat (Zhang et al., 1991) and sheep (Cozzi et al., 1992) pineal gland most of the tyrosine hydroxylase (TH; the rate-limiting enzyme for NE synthesis) immunoreactive fibers disappear after
