Melatonin in type 2 diabetes mellitus and obesity

Article metrics

Abstract

Despite considerable advances in the past few years, obesity and type 2 diabetes mellitus (T2DM) remain two major challenges for public health systems globally. In the past 9 years, genome-wide association studies (GWAS) have established a major role for genetic variation within the MTNR1B locus in regulating fasting plasma levels of glucose and in affecting the risk of T2DM. This discovery generated a major interest in the melatonergic system, in particular the melatonin MT2 receptor (which is encoded by MTNR1B). In this Review, we discuss the effect of melatonin and its receptors on glucose homeostasis, obesity and T2DM. Preclinical and clinical post-GWAS evidence of frequent and rare variants of the MTNR1B locus confirmed its importance in regulating glucose homeostasis and T2DM risk with minor effects on obesity. However, these studies did not solve the question of whether melatonin is beneficial or detrimental, an issue that will be discussed in the context of the peculiarities of the melatonergic system. Melatonin receptors might have therapeutic potential as they belong to the highly druggable G protein-coupled receptor superfamily. Clarifying the precise role of melatonin and its receptors on glucose homeostasis is urgent, as melatonin is widely used for other indications, either as a prescribed medication or as a supplement without medical prescription, in many countries in Europe and in the USA.

Key points

  • The rs10830963 single-nucleotide polymorphism (SNP) in the MTNR1B locus is associated with increased fasting plasma glucose levels and impaired insulin secretion, as well as increased risk of type 2 diabetes mellitus (T2DM) and gestational diabetes mellitus.

  • Obesity seems to not be associated with the rs10830963 SNP in adults but might have a role in fetal birth weight.

  • The phenotype of rs10830963 risk allele carriers includes increased MTNR1B mRNA expression, altered melatonin secretion and possibly further effects associated with the enhancer activity of the region surrounding the rs10830963 SNP.

  • Loss-of-function of rare MT2 receptor variants, in particular of melatonin-induced Gi1 and Gz and spontaneous β-arrestin 2 recruitment, is associated with increased risk of T2DM.

  • Lifestyle recommendations are emerging for rs10830963 risk allele carriers and further clinical evidence has to be gathered to evaluate the prescription of melatonin for patients with T2DM.

  • The wide use of melatonin by millions of people, both as a supplement and as a medicine, calls for a rapid assessment of the effect of melatonin on glucose homeostasis.

Introduction

Melatonin is a pleiotropic hormone primarily known for its regulatory role in circadian and seasonal rhythms, sleep, retinal functions and the immune system1. Studies published in the past 9 years established a major role for the MTNR1B locus, which encodes the melatonin MT2 receptor (also known as melatonin receptor type 1B), in regulating fasting plasma glucose (FPG) levels and the risk of type 2 diabetes mellitus (T2DM)2,3,4. This discovery generated a lot of interest in the metabolic effects of melatonin, particularly in its role in the development of T2DM and obesity. As a member of the G protein-coupled receptor (GPCR) family, the MT2 receptor has a high potential for drug development, as a target for new therapeutics. Indeed, melatonin is already widely used for several non-metabolic indications and is also available over the counter in several countries, thus highlighting the need to better understand its metabolic effects. This Review will discuss the effect of melatonin and melatonin receptor signalling on glucose homeostasis and energy metabolism at the cellular, tissue and individual organism level.

Melatonin is unique as it can mediate different types of effects. These effects include immediate effects that occur during the night and are determined by the nocturnal secretion pattern of melatonin, prospective or delayed effects that are typically primed during the night with functional consequences during the day, chronobiotic effects that rely on the direct effect of melatonin on the circadian clock and seasonal effects that depend on the duration of the night. For instance, melatonin is involved in the regulation of sleep in diurnal species (including humans), the transition of night length information to the suprachiasmatic nuclei (SCN) and other organs5 and the temporal organization and circadian distribution of several metabolic processes associated with energy balance and seasonal control of photoperiodic functions such as reproduction6.

This Review will mainly focus on the immediate and delayed effects, whereas chronobiotic and seasonal effects of melatonin, although likely to contribute to the metabolic effects of melatonin, will be only briefly discussed. The contribution of melatonin to the development of obesity and T2DM and the potential of targeting the melatonin system for the treatment of these diseases will be discussed. A clear distinction will be made between animal models (that is, mice and rats) and data obtained in humans. The major contribution of genetics in identifying frequent and rare variants of the MTNR1B locus, including the current state of the functional characterization of the genetic variants in this gene that are associated with the risk of T2DM, will be discussed. The resulting controversial findings regarding the effect of various natural variants in the MTNR1B locus on metabolic traits related to T2DM will be critically analysed and discussed in the context of the peculiarities of the melatonin system. In the last part of the Review, lifestyle recommendations and pharmacological targeting of the melatonin system to manage the risk of T2DM will be put into perspective.

Melatonin synthesis and signalling

Melatonin synthesis

Melatonin is an old molecule in evolutionary terms that seems to have appeared 2.5–3.0 billion years ago in photosynthetic cyanobacteria, probably fulfilling the function of an antioxidant7. Subsequently, melatonin became the key molecule conveying the environmental light/dark information and a ligand for membrane receptors to mediate actions on a variety of additional intracellular processes. The conserved biosynthetic pathway of melatonin starts with the hydroxylation of tryptophan to 5-hydroxytryptophan, followed by decarboxylation to generate serotonin. Serotonin is then acetylated by the rate-limiting enzyme arylalkylamine N-acetyltransferase to give N-acetylserotonin, which is finally converted by the acetyl-serotonin-methyltransferase to melatonin by O-methylation8. In vertebrates, the pineal gland is the main source of melatonin production. Following its synthesis, melatonin is immediately released into the cerebrospinal fluid (CSF) and bloodstream9,10, with the melatonin content of the CSF reaching levels 100-fold higher than that in peripheral structures11. Importantly, melatonin from the pineal gland is synthesized in a circadian manner, under the control of the central biological clock — the SCN — with the highest circulating levels at night and low values during the day12. Thus, the circadian regulation of the central biological clock by melatonin, as a feedback mechanism, is reliant on the rhythmic melatonin concentrations in the CSF, whereas its rhythmic concentrations in the blood influences the circadian regulation of peripheral tissues; both functionalities contribute to whole body synchronization as a result of the chronobiotic properties of melatonin13. Melatonin is rapidly metabolized in the body, with a half-life of 20–30 minutes14.

Local melatonin synthesis has been reported in other tissues and cells in mammals, which has a minor effect on plasma or CSF levels of melatonin. For instance, melatonin is synthesized in the retina in a circadian manner and participates in the regulation of retinal physiology, including retinal light sensitivity during the night15. Local, non-rhythmic melatonin synthesis has also been reported in activated macrophages and lymphocytes16,17, where melatonin is reported to participate in the inflammatory response of the organism via promotion of M2-macrophage polarization18, enhancement of the phagocytic activity of macrophages and regulation of the production of cytokines that participate in the resolution of inflammation17. The gastrointestinal epithelium also produces melatonin from its precursor serotonin, which is thought to come from enterochromaffin cells19. In the gut, melatonin seems to participate in the regulation of intestinal motility, immune system responses, ion transportation in the lower gut and the release of peptides involved in energy balance such as peptide YY20. Whereas melatonin synthesis is classically considered to take place in the cytoplasm, in the past 5 years, non-rhythmic melatonin synthesis has been reported to occur in isolated mitochondria from mouse brain cells21 and oocytes22, as well as in plant mitochondria and chloroplasts23. The full meaning of this intriguing finding remains to be clarified, including the question of the effect of mitochondrial melatonin synthesis in comparison to pineal melatonin synthesis.

Signalling and metabolic functions

Melatonin has been reported to bind with high affinity to MT1 and MT2 receptors1. These receptors are members of the GPCR superfamily and primarily couple to G proteins of the Gi/o and Gq/11 subclasses24, which trigger multiple downstream signalling pathways. Both receptors also couple to β-arrestins, the second major mediator of GPCR function, but the consequences of this coupling are unknown25,26,27. By combining genetic and proteomic approaches, the interactome of melatonin receptors was estimated at around 200 interactors for each receptor28. Many of these interactors are involved in the regulation of receptor function29,30. Of the multiple signalling pathways activated by melatonin receptors, most have been reported to be dependent on pertussis toxin-sensitive Gi/o proteins, mediating the inhibition of the adenylyl cyclase–protein kinase A (PKA)–cAMP-responsive element binding (CREB) pathway, the soluble guanylyl cyclase (sGC)–protein kinase G (PKG) pathway and the activity of various kinases (such as phosphatidylinositol 3-kinase (PI3K), AKT, protein kinase C (PKC) and extracellular-signal-regulated kinase (ERK)) and ion channels (such as large-conductance Ca2+-activated K+ BKCa and Kir3 channel)31. In several other cases, activation of the Gq/11 protein–PKC–phospholipase C (PLC) pathway has been reported in diverse physiological settings32. Some reports suggest that Gs and G16 proteins are mediators of the melatonin signal, however, the physiological relevance of these observations remains to be demonstrated33,34.

Melatonin receptors are expressed at central and peripheral sites, which can both contribute to the regulation of metabolic functions. The effect of central melatonin receptors on metabolism is probably mediated through the hypothalamic SCN, which is the biological master clock. Indeed, melatonin is known to entrain the circadian rhythm of the SCN and is currently the only approved pharmacological tool to treat circadian dysfunction of the SCN in humans35. Studies published in the past few years indicate that the circadian system is important in regulating the daily rhythm of plasma metabolites36 and glucose metabolism37 and that disruption of the master clock leads to metabolic diseases such as T2DM38 (Box 1). Therefore, alteration of the melatonin system has a direct effect on the function of the master clock in the SCN, which could affect the development of metabolic diseases. In the SCN, melatonin inhibits the cAMP–PKA–CREB pathway39 and activates the PKC pathway40,41 and Kir3 channels42,43 to acutely inhibit the neuron firing rate, to phase-shift the circadian rhythm of neuronal firing and to regulate clock gene expression44,45,46,47 (Fig. 1).

Fig. 1: Metabolic processes influenced by melatonin signalling in peripheral and central tissues.
figure1

Melatonin receptors are expressed at central and peripheral sites at which they contribute to the regulation of metabolic functions. Melatonin is secreted by the pineal gland in a circadian manner under the control of the master clock in the hypothalamic suprachiasmatic nucleus (SCN). Melatonin receptors in the SCN in turn inhibit the cAMP–protein kinase A (PKA)–cAMP-responsive element binding (CREB) pathway and activate the protein kinase C (PKC) pathway and Kir3 channels to regulate acute and circadian neuronal firing and clock gene expression. Melatonin in the circulation then modulates metabolic processes in the periphery either by directly acting on peripheral organs or indirectly by modulating the circadian activity of the central master clock. In mouse liver, melatonin is required for insulin-stimulated phosphatidylinositol 3-kinase (PI3K)–AKT activity, in rats it suppresses hepatic glucose production and in the human hepatocyte cell line HepG2 it activates glycogen synthesis, probably via a PKCζ–AKT–glycogen synthase kinase-3β (GSK3β) pathway. In mouse skeletal muscle, melatonin activates the insulin receptor substrate 1 (IRS1)–PI3K–PKCζ pathway to enhance the rate of glucose uptake. In inguinal rat adipocytes, melatonin inhibits the cAMP–PKA pathway and isoproterenol-induced lipolysis and fatty acid transport in some cases. In the human brown adipocyte PAZ6 cell line, melatonin acutely inhibits cGMP production and decreases glucose transporter type 4 (GLUT4) expression and glucose uptake upon long-term treatment. In rodent pancreatic β-cells, melatonin acutely reduces insulin release through inhibition of cAMP and cGMP levels, whereas it sensitizes the cAMP pathway and promotes insulin secretion upon long-term (physiological) melatonin stimulation. In human pancreatic islets, melatonin treatment increases insulin secretion and promotes β-cell survival via decreased JUN N-terminal kinase (JNK) activation. Melatonin administration in mouse pancreatic α-cells and in human pancreatic islets increases glucagon secretion, possibly by activating the Gq/11–PLC–PI3K cascade. Glucagon, in turn, can promote insulin secretion from β-cells and somatostatin secretion in δ-cells via its Gs-coupled glucagon receptor (GCGR). Finally, in human pancreatic δ-cells, melatonin has been reported to influence somatostatin secretion in both ways via modulation of cAMP levels. Broken lines correspond to suggested functions of melatonin receptors for which the precise pathway has to be further elucidated. The dark blue wave-formed lines illustrate circadian clock oscillations.

Expression of melatonin receptors in the periphery occurs in many tissues, but at low levels. Despite the fact that numerous animal studies suggest that melatonin has a modulatory effect on peripheral tissues, a direct effect of melatonin on these tissues has only been demonstrated in a limited number of studies47,48,49,50,51,52,53. Similar to the SCN, regulation of glucose metabolism by melatonin in peripheral tissues might involve the regulation of clock genes at the local level, but the contribution of the peripheral versus central receptors and the corresponding signal transduction pathways remain unknown. A direct effect of melatonin on specific tissue functions, such as glucose uptake or insulin secretion, has been reported. For instance, in the human hepatocyte cell line HepG2, melatonin activates the function of PKCζ, AKT and glycogen synthase kinase-3β (GSK3β) to stimulate glycogen synthesis54; melatonin also reduces hepatic gluconeogenesis in rats48. Analysis of both MT1-knockout and MT2-knockout mice suggests an important role for melatonin signalling in the regulation of hepatic insulin-dependent glucose metabolism via augmentation of PI3K–AKT activity55,56. In mouse skeletal muscle cells, melatonin activates the insulin receptor substrate 1 (IRS1)–PI3K–PKCζ pathway and stimulates glucose transport47, a finding that is in agreement with subsequent findings of reduced glucose uptake in skeletal muscle of MT1-knockout mice55. In isolated inguinal rat adipocytes, melatonin inhibits isoproterenol-induced lipolysis and fatty acid transport through inhibition of the cAMP–PKA pathway49,51. In the human brown adipocyte PAZ6 cell line, acute melatonin treatment inhibits cAMP and cGMP production; long-term treatment decreases glucose transporter type 4 (GLUT4) expression and glucose uptake, thus participating in the delayed effects of melatonin57 (Fig. 1).

An increasing number of studies have focused on the role of melatonin in pancreatic islet populations (Box 2) and the effect of melatonin signalling on their function. In rodent β-cells, the predominant effect of melatonin is the reduction of insulin release through inhibition of the Gi–cAMP–PKA50,58,59 or cGMP52 pathways, while a marked increase in insulin release was observed upon long-term treatment of the insulinoma cell line INS-1 with melatonin53,58,59. In the murine pancreatic α-cell line αTC1.9, melatonin administration increased glucagon secretion, possibly via Gq/11-dependent activation of the PLC–PI3K cascade60. Studies on human islets suggest that melatonin increases intracellular levels of calcium and stimulates glucagon and insulin release from α-cells and β-cells, respectively61, probably via paracrine input from α-cells to insulin-secreting β-cells. Importantly, similar to rodent cells, long-term activation of melatonin signalling (of a duration that mimics night exposure) in human islets using melatonin or the melatonin receptor agonist ramelteon increases islet sensitivity to cAMP, which results in increased insulin secretion62. Insulin secretion and β-cell survival are improved in response to melatonin signalling, as shown by the decreased proteotoxicity-induced cell apoptosis and oxidative stress in human islets exposed to chronic hyperglycaemia and in islets from patients with T2DM62, as well as by the preservation of the protein levels of the histone acetyl transferase p300 (ref.63). These findings are in agreement with previous reports on the role of melatonin in preventing β-cell damage58. Melatonin administration in the human pancreatic δ-cell line QGP-1 decreases somatostatin secretion via mechanisms involving decreased cAMP concentrations64. By contrast, exposure of human islets to melatonin increases somatostatin release65, which highlights an influence of melatonin in δ-cells and the need for further experimentation.

Collectively, these data show that melatonin can act through central and peripheral receptors with well-documented regulatory effects on the central circadian master clock and less well-documented effects on the function of peripheral tissues. Whereas the effect of melatonin on lipid metabolism does not seem to be predominant, its involvement in glucose homeostasis, in particular on glucose uptake, pancreatic insulin secretion and β-cell survival, has been extensively observed. Acute versus long-term stimulation paradigms seem to be important for the functional outcome (reduced versus increased insulin secretion), which reflects the different outcomes of acute versus delayed effects of melatonin. Paracrine interactions between different pancreatic cell types have to be considered before conclusions are drawn on the effect of melatonin on insulin secretion. In addition, important species-specific differences might exist between rodent models and humans, which might have an influence on the effect of melatonin on islet function and glucose homeostasis.

Melatonin in metabolic diseases

Melatonin regulates a variety of physiological and neuroendocrine functions in mammals, such as circadian rhythms, sleep onset and architecture, retinal function and immune function, as well as seasonal reproduction. The critical role of melatonin in the photoperiodic control of seasonal reproduction has been demonstrated by use of artificial patterns of timed melatonin infusion66 and in pinealectomised animals67. These initial findings on the effects on seasonal reproduction, together with findings from the past 9 years, contributed to the increasing recognition that melatonin is an important modulator of metabolic functions and associated diseases6,24,58,68. Here we will discuss the effect of melatonin on the development of obesity and T2DM.

Melatonin in obesity

Hibernating animals are an interesting model of body weight regulation as they undergo dramatic body weight changes during the annual cycle with massive weight gain when in preparation for hibernation, followed by weight loss during hibernation. According to the seasonal effects of melatonin, the duration of melatonin synthesis during the night reflects the variation of the day length during different seasons. Importantly, in seasonal animals such as Siberian hamsters this variation in melatonin signal triggers not only the reproductive cycle, but also a massive increase in body weight in anticipation of the reduced food intake during hibernation69. In this model, melatonin modulates the sympathetic nervous system through central melatonin receptors and acts directly on the adipose tissue70. As melatonin acts on mitochondrial gene expression in the brown adipocytes of Siberian hamster, it is plausible that the effects of melatonin are mediated via modulation of mitochondrial gene expression71. In rats and mice, melatonin supplementation decreases body weight gain and the amount of adipose tissue in several experimental models6. These effects are probably mediated by central and peripheral melatonin target tissues, which results in synchronization of circadian rhythms to the activity-feeding–rest-fasting cycle and amelioration of glucose uptake directly on adipocytes57,72. Even so, both MT1-knockout and MT2-knockout mice display insulin resistance, which suggests that the receptors are implicated in the regulation of other metabolic processes in nocturnal species rather than regulation of body weight55,56,73.

Reports on the beneficial effects of melatonin on body weight in animal models, together with the absence of toxicity of melatonin, prompted several small-scale studies in humans (<100 participants) to evaluate the effect of melatonin administration alone or in combination with other medications in either individuals with obesity or patients with metabolic disorders (Table 1). In treatment protocols varying from 1 day up to 1 year, a modest reduction in body weight and plasma levels of lipids was observed in some studies, but not in others, suggesting an overall modest, if any, effect of melatonin (see Table 1 for references).

Table 1 Human studies on the influence of melatonin or melatonin receptor agonists

Genetic studies have also evaluated the influence of the melatonergic system on body weight regulation, obesity and lipid metabolism. Neither genome-wide association studies (GWAS) nor most of the more targeted studies focusing on MTNR1A and MTNR1B, which encode the MT1 and MT2 receptors, respectively, detected a consistent association of genes of the melatonergic system with risk of obesity74,75,76,77,78. The rs8192552 (Gly24Glu) variant, located in the coding region of the MTNR1B gene, was reported to be functionally defective and associated with increased BMI79; however, this finding was not replicated in other studies75,80.

GWAS focusing on the role of maternal genetic variation on birth weight revealed an association of rs10830963 and rs1387153 variants, located in the MTNR1B locus, with increased birth weight81. This finding highlights the effect of the maternal genotype and the intrauterine environment on the fetus. Interestingly, carriers of the rs10830963 risk allele show extended melatonin secretion in the morning82. As maternal melatonin is known to cross the placenta83 and to affect fetal growth and maturation84, it is tempting to speculate that the modified melatonin secretion pattern of the mother might affect the fetus by prolonging the immediate effects or by altering the chronobiotic effects of melatonin.

Several studies have started to examine the effect of the rs10830963 variant, which is known to be associated with increased FPG levels and risk of T2DM2,3,4 (see subsequent sections for more details), on body weight regulation in dietary weight loss intervention protocols (Table 2). Compared with carriers of non-risk allele, carriers of the risk allele showed a greater improvement in body adiposity and fat distribution and reduction in cholesterol levels but only when eating a low-fat diet, and not under high-fat diet conditions77,78,85,86. In other studies, rs10830963 risk allele carriers did not benefit from weight loss programmes, or showed only modest improvements87,88.

Table 2 Human interventional studies testing the effect of the rs10830963 SNP

Taken together, the available data do not support a major role of melatonin in body weight regulation and lipid metabolism in adults, but maternal melatonin regulation might have a notable effect on fetal birth weight. Intervention studies are still limited by the low number of participants, the variability of the applied protocols, the high number of confounding environmental parameters and the lack of information on melatonin secretion profiles in most cases.

Melatonin in T2DM

Studies in rodents indicate a general beneficial role of melatonin on glucose homeostasis. Absence of melatonin in pinealectomized animals leads to a reduction in GLUT4 gene expression and protein content, glucose intolerance and peripheral and central insulin resistance, which are reversed by melatonin treatment89,90. In rats with streptozotocin-induced T2DM, combined treatment with melatonin and insulin promoted better glycaemic control and improved insulin sensitivity in white adipose tissue compared with the insulin or melatonin treatment alone91. Decreased melatonin synthesis has been reported in several animal models of T2DM92,93 and improvement of glucose metabolism after melatonin administration was observed in the high-fat diet-fed insulin-resistant mouse model94.

The role of melatonin in glucose homeostasis in rodents was further studied in melatonin receptor knockout mice (Box 3). MT1-knockout mice show a robust metabolic phenotype with systemic insulin resistance, which is probably the result of decreased insulin sensitivity and a pronounced inability of insulin to suppress hepatic glucose production55,73. MT2-knockout mice also exhibit reduced hepatic insulin sensitivity but, in contrast to MT1-knockout mice, show increased insulin release56. In humans, nocturnal melatonin levels have been reported to be lower in patients with T2DM than in controls93,95. Several small-scale controlled clinical studies showed that a single acute melatonin treatment worsens morning and evening glucose tolerance, both in older, healthy postmenopausal women (52 ± 6 years of age)96 and younger healthy women (24 ± 6 years of age)97. By contrast, repeated administration over a 5-month period tends to have beneficial effects; HbA1c levels decreased, which suggests improved glycaemic control98. Combined administration of melatonin and zinc acetate with metformin in patients with poorly controlled T2DM improved the tissue responses to metformin99. Similarly, the median value of insulin resistance as defined by HOMA was statistically significantly reduced in patients with nonalcoholic steatohepatitis who were treated with melatonin for 1 month100.

Frequent variants of the MTNR1B locus

Among the candidate loci associated with increased risk of T2DM that have been identified in GWAS is the MTNR1B locus (reported in 2009 by three independent research consortia in cohorts of European origin). An association was described between the rs1387153 single-nucleotide polymorphism (SNP) (located 29 kb upstream of exon 1 of the MTNR1B gene) and increased FPG levels and T2DM risk2. Similar associations with stronger signals were reported for the rs10830963 SNP (located in the intronic region of the MTNR1B gene), which was in complete linkage disequilibrium with rs1387153 (ref.4). Likewise, Lyssenko and colleagues reported an association between the rs10830963 SNP and impaired insulin secretion3, which suggests that MT2 action is focused in the pancreas as a target tissue of melatonin. The effect of the rs10830963 SNP on increased FPG levels or β-cell function was replicated in other European cohorts101,102,103,104,105 and by studies involving different ethnic groups, including Japanese106,107, African American108,109, Mexican American110, Chinese111,112,113,114,115 and Sri Lankan115 people, whereas moderate effects were reported for Asian Indian people116,117,118. Importantly, the effect on FPG was also replicated in healthy children and adolescents of European119,120, Asian121 and Mexican122 origin, as well as in children and adolescents with obesity123,124,125, suggesting that the effect of the rs10830963 risk allele is apparent at early ages.

The MTNR1B locus was also associated with poor glycaemia over a 2–3 month period, as detected through a GWAS analysis of variants associated with increased levels of HbA1c (ref.126). Furthermore, one of the strong association signals observed for rs10830963 SNP carriers is defective early-phase insulin release, which might contribute to the increased plasma levels of glucose observed in the carriers who did not have T2DM127,128,129,130, a finding that was replicated in two meta-analysis studies published in the past 5 years131,132. Moreover, the rs10830963 SNP was shown to effect the rate of progression from normal fasting glucose to impaired fasting glucose, but not the rate of progression from impaired fasting glucose to T2DM, indicating that the variant might also be important for the development of prediabetic fasting hyperglycaemia with a minor effect on the transition to the diabetic state133. Only a few studies have reported an association between the rs10830963 SNP and insulin resistance101,134. In a study published in 2018 that proposed an extended classification of patients with diabetes mellitus into six subgroups based on a cluster analysis of available metabolic, genetic and clinical data, the rs10830963 SNP was surprisingly associated most significantly (P = 0.05) with cluster 1 (labelled as severe autoimmune diabetes mellitus that overlaps with type 1 diabetes mellitus) and not with cluster 2 (characterized by severe insulin-deficient diabetes mellitus), as would have been expected from previous data135.

The rs10830963 SNP has also been reported in many association studies as a risk factor for gestational diabetes mellitus (GDM) in different ethnic groups136,137,138,139,140,141,142. Interestingly, the association with GDM was restricted to those with increased pre-pregnancy BMI (≥25 kg/m2), indicating that obesity might also have a role in GDM risk142. Considering that GDM itself is a strong risk factor for developing T2DM later in life and influences the metabolic health of the offspring143, identification of genetic risk factors is required in determining lifestyle interventions with a preventive or therapeutic goal. Taken together, the association of the rs10830963 SNP with FPG and T2DM risk has been extensively replicated. The involvement of rs10830963 in the development of prediabetic fasting hyperglycaemia effects insulin secretion and has a minor effect on liver insulin sensitivity.

Rare MTNR1B variants: rare but insightful

Many frequent variants associated with diseases are located in genomic regions with unknown function. By contrast, rare variants, which are typically discovered by targeted sequencing of the flanking or coding region of genes of interest, lead to more straightforward and testable hypotheses concerning their functional effect. Following this idea, a causal link between the MTNR1B locus and T2DM risk has been illustrated for the first time by a large-scale exon re-sequencing study of the MTNR1B gene in combination with systematic functional characterization of each corresponding MT2 mutant75. A study that sequenced >7,600 Europeans revealed 40 non-synonymous variants (Fig. 2) of which the 36 very rare ones (minor allele frequency (MAF) below 0.1%) were significantly associated with T2DM, while the remaining four (frequent or rare ones with a MAF higher than 0.1%) did not contribute to T2DM risk. Functional characterization of the MT2 mutants in terms of melatonin binding, cell surface expression and cAMP signalling revealed that four mutants completely lost their capacity to bind melatonin and an additional nine mutants did not inhibit cAMP production, despite normal cell surface expression. Statistical analysis showed that only mutants with a loss-of-function phenotype strongly associated with increased T2DM risk (OR of 5, when analyzed in aggregation)75, establishing a firm functional link between loss of MT2 receptor function and T2DM.

Fig. 2: Melatonin MT2 receptor mutants and signalling defects associated with type 2 diabetes mellitus.
figure2

MT2 activates the function of Gi1 and Gz proteins, inhibits cAMP production, promotes extracellular-signal-regulated kinases 1 and 2 (ERK1/2) phosphorylation and recruits β-arrestin 2 in a spontaneous and melatonin-induced manner. MT2 mutations are shown on the receptor sequence. Those in red result in defects in at least one of the aforementioned pathways. Defective activation of melatonin-induced Gi1 and Gz proteins and spontaneous β-arrestin 2 recruitment to MT2 are independently associated with the risk of type 2 diabetes mellitus. Defective MT2 mutants are listed for each of these parameters. The blue lines correspond to the wild-type MT2 and red lines to the variant MT2.

Signalling pathway-specific defects of rare MT2 mutants

GPCRs such as MT2 might engage multiple distinct signalling pathways. Amplifying or restoring the activation of pathways associated with the therapeutically desired effect with pathway-biased orthosteric or allosteric ligands, while excluding pathways associated with undesirable adverse effects, is an interesting therapeutic goal. Defining the functional defect of receptor mutants is thus of therapeutic relevance for the establishment of tailored medical interventions for risk allele carriers. As a first step toward this goal we assessed the effect of all 40 MT2 mutants on multiple pathways to investigate if the association between MT2 loss-of-function and T2DM is linked to specific signalling defects80 (Fig. 2). The study tested both spontaneous (ligand-independent) and melatonin-induced MT2 activity for five signalling events; namely Gi1 and Gz activation, inhibition of cAMP production, β-arrestin 2 recruitment and ERK1/2 activation. Only ten MT2 mutants had a phenotype that was indistinguishable from the wild-type, whereas the rest showed functional defects in one or more of the tested functional parameters (Table 3). Detailed genetic association analysis detected three parameters that stand out in terms of significance of association with T2DM risk. These include 19 MT2 mutants with defects in melatonin-induced Gi1 activation, 22 MT2 mutants with defects in melatonin-induced Gz activation and 15 MT2 mutants with defective spontaneous (ligand-independent) β-arrestin 2 recruitment, while defects in other signalling parameters showed only a trend of association with T2DM risk. Importantly, as previously noted75, none of the ‘neutral’ rare variants were associated with T2DM risk.

Table 3 Defects in functional parameters of rare MT2 mutants

The reported results not only introduce Gz and β-arrestin 2 recruitment as new players in MT2 signalling that have a potential effect on T2DM risk, but also put spontaneous signalling of MT2 at the forefront as a risk factor for T2DM. Whereas the capacity of the mutants to activate Gz and Gi1 proteins correlated nicely, G protein activation correlated only poorly with β-arrestin 2 recruitment, which indicates that the influence of G protein-dependent effects on T2DM needs to be addressed independently of the effect of β-arrestin 2 recruitment. Concerning the activation of Gz by MT2 and the defective coupling with Gz by certain mutants, it is important to note that Gz expression is restricted to certain brain regions and retinal ganglion cells144, areas where MT2 is expressed15,145, which suggests that the tissue-specific context should also be considered when evaluating the effect of receptor dysfunction. Collectively, this large-scale functional study revealed associations of T2DM risk with defects of specific melatonin signalling pathways, namely melatonin-induced Gi1 and Gz and spontaneous β-arrestin 2 recruitment, opening avenues for pathway-specific personalized therapeutic interventions.

Controversial issues

Even in light of all available data, the role of melatonin in glucose homeostasis remains complex and the evidence is sometimes contradictory. The most prevailing current hypothesis is based on the action of melatonin on impairing glucose homeostasis, which would be mediated by its inhibitory action on insulin secretion. A key finding for this hypothesis comes from the observation that MTNR1B mRNA expression levels are ~4 times higher in human pancreatic islets of carriers of the common rs10830963 variant (a risk allele) than in islets of people without this variant3. Accordingly, increased melatonin action is expected to increase melatonin signalling in islets and reduce insulin secretion, leading to hyperglycaemia and an increased risk of T2DM56,146. At first glance, this model seems to be simple and compatible with the phenotype of rs10830963 risk allele carriers, which comprises increased FPG levels and T2DM risk3. Consequently, limiting the action of the melatonergic system (that is, by using melatonin receptor antagonists) would be the advised treatment for patients with T2DM who have the risk allele. However, the situation is not as clear as that, as the reduced night-time melatonin levels observed in patients with T2DM93,95 and the multiple rare MT2 mutants that result in a loss-of-function phenotype and are associated with increased T2DM risk75,80 (Fig. 2; Table 3) suggest rather the opposite, namely that dampening of the melatonin system leads to the development of T2DM75,147.

These controversial statements have been discussed in several commentaries and reviews146,147,148,149 but an explanation is still missing. Some have tried to reconcile the results by proposing an ‘equilibrium hypothesis’ where diversion from the equilibrium (normal situation), either by exaggerated or dampened melatonin function in frequent and rare variant carriers, respectively, becomes detrimental for glucose homeostasis146. Others suggested an ‘age-related chronobiological hypothesis’ that emphasizes the importance of the circadian system and its ageing-related deterioration in the understanding of melatonin’s actions150. Still others defended the ‘loss-of-function hypothesis’ by questioning the relevance of pancreatic MT2 receptors and by postulating a predominant indirect effect of melatonin on glucose metabolism through its well-established regulatory role on the central biological clock in the SCN149. None of these hypotheses are currently satisfying, or they need more experimental evidence to become generally accepted. Further adding to the complexity, the issue of insulin sensitization in the progression from prediabetes to T2DM should be considered to understand the effect of melatonin on insulin release in the early stages of T2DM. Revisiting some specific features of the melatonin system, including the different types of effects that melatonin has and the regulation (desensitization and sensitization) of melatonin responses, might help to understand how this hormone functions and improve comprehension of the apparent contradictions to pave the way towards a unifying model that can be agreed on by all in the field.

Beyond rodent models

Mice and rats are useful models for physiological studies in mammals that offer the possibility of targeted deletion of genes of interest. However, these nocturnal rodents might be of limited use to understand the relationship between melatonin and the risk of T2DM in humans, as the night-time melatonin peak coincides with the eating phase in nocturnal rodents and with the resting and sleep phase in diurnal humans, illustrating their fundamentally different needs in regulating glucose levels during the night. Furthermore, different diurnal fluctuations of glucose tolerance and insulin sensitivity exist in rodents and humans151. In humans, both are highest during the early morning with a progressive decline towards sleep onset. In nocturnal rodents, glucose tolerance and insulin sensitivity are in phase opposition to humans illustrating the fundamental difference in decoding the melatonin message in diurnal and nocturnal species.

Melatonin regulates multiple functions

Melatonin is a pleiotropic hormone that regulates multiple physiological functions, a feature that is expected for a molecule that is considered to be the signal informing multiple tissues that night is beginning. Surprisingly, when it comes to explaining the action of melatonin on glucose homeostasis, the inhibitory action of melatonin on insulin secretion in pancreatic β-cells is often put at the forefront. This emphasis is mainly based on the well-known coupling of melatonin receptors to the Gi–cAMP pathway1,24 and genetic association studies involving the rs10830963 SNP (see previous section). However, most of the functional studies on insulin release have been obtained in rodents and several similar studies in human β-cell lines and islets suggest rather the opposite, that is, a stimulatory role for melatonin on insulin secretion. This discrepancy might be explained by differences in the stimulation protocol (short-term versus long-term stimulation)53,62, as the cAMP system is known to become sensitized upon long-term melatonin stimulation152. Another explanation might rely on an indirect, stimulatory effect of melatonin on glucagon secretion by pancreatic α-cells through a Gq/11–PLC-dependent pathway and subsequent insulin secretion by β-cells through Gs-coupled glucagon receptors60,61,153 (Fig. 1). Stimulation of glucagon secretion by melatonin might be further potentiated by food intake during the biological night as glucagon upregulation presents one of the main modifications observed under conditions of circadian misalignment in humans154.

The opposing effects of melatonin on insulin secretion through Gi/o and Gq/11 proteins are also an integral part of the ‘equilibrium hypothesis’, which seeks to reconcile the controversial findings on common and rare MT2 variants. However, in our opinion the biological relevance of this Gi/o to Gq/11 balance remains to be demonstrated. As a matter of fact, melatonin modulates many other cellular functions in pancreatic islets and other cell types31, such as cell survival62,63,155, cell proliferation156,157 and release of Ca2+ (ref.153) and cytochrome c21, that might be as relevant for glucose homeostasis as the direct regulation of insulin secretion (Fig. 1).

Beyond the Gi/o–cAMP pathway

Much attention has also been given to the inhibitory effect of melatonin on the Gi/o–cAMP pathway, which is the most extensively studied signalling pathway of melatonin receptors but is certainly not the only signalling pathway for these receptors. Further downstream effectors of Gi/o activation, such as ion channels, kinases (PI3K, AKT and ERK1/2) and sGC have been reported (see previous section), and are of potential relevance for the regulation of glucose homeostasis by melatonin31. In addition to Gi activation, melatonin receptors activate the Gq/11–PLC pathway153 and recruit β-arrestins (Fig. 1). Further interesting new insights are coming from the signalling profiles of rare MT2 receptor variants that have been associated with increased risk of T2DM80. Indeed, in addition to Gi/o protein activation, defective Gz protein activation and β-arrestin 2 recruitment are associated with increased risk of T2DM80 (Fig. 2). These findings provide two new lines of investigation that have to be considered to fully understand the association of MT2 receptor function with T2DM risk, possibly not only in pancreatic β-cells but also in the brain144,158.

Low levels of melatonin receptor expression

Melatonin receptor expression is widespread but at low to very low levels (~1 fmol/mg of protein or lower)28,31. Unfortunately, after extensive validation we and many other laboratories have concluded that commercially available antibodies against melatonin receptors are not reliable enough to detect endogenous proteins in rodents159 and, apart from some exceptions, are even unable to detect the recombinant receptors160. Considerable expression of melatonin receptors is only observed at three sites — the retina, the hypothalamic SCN and the hypophysal pars tuberalis24. At most other sites, melatonin receptor expression remains controversial and most of the evidence for its existence relies on functional studies.

However, absence of detection of receptor expression does not mean that these receptors are not functionally relevant, as demonstrated for the expression of the MT2 receptor, which is barely detectable by in situ hybridization and 2-[125I]-iodomelatonin binding in the SCN but has an important role in the regulation of the biological master clock at this site161. Not surprisingly, expression of melatonin receptors in peripheral tissues related to glucose metabolism in humans is similarly problematic and controversial. Expression in insulin-sensitive tissues in humans has only been reported in brown and white adipose tissue57 and pancreas at the mRNA level2. Since the discovery of the common rs10830963 variant that is associated with T2DM risk and the reported increase in MTNR1B mRNA levels in carriers of the risk allele3, expression of melatonin receptors in human pancreatic islets was intensely studied using various techniques, such as expression quantitative trait loci (eQTL) studies56,162,163 and single-cell RNA-sequencing164,165. All studies agree that melatonin receptor mRNA levels are low, in some instances even absent or only detected in a subset of cells. In conclusion, in our opinion, unfortunately MT2 expression studies are unlikely to provide convincing supportive evidence for the functional relevance of MT2 in the human pancreas due to the expression of very low mRNA levels and the lack of satisfactory tools (that is, antibodies) to detect the MT2 protein.

Increased MT2 mRNA expression

Increased MT2 expression could be a relevant feature of rs10830963 risk allele carriers that might explain their T2DM phenotype; however, currently available data do not support this conclusion. Whether the observed difference in MTNR1B mRNA levels also translates into increased MT2 expression levels remains unknown. In addition, an increase in receptor number does not necessarily result in an increased receptor signalling capacity due to spare receptors or chronic receptor desensitization, as shown for MT2 in in vitro experiments80,166. Beyond the questions of whether variation of MT2 expression contributes to the T2DM phenotype of rs10830963 risk allele carriers, other modifications cannot be excluded and have to be taken into account, as illustrated by the significantly longer (~10%) duration of melatonin secretion that has been reported in rs10830963 risk allele carriers compared with non-carriers82. Intriguingly, increased MTNR1B mRNA levels have not been reported in pancreatic islets of donors with T2DM versus normoglycaemic donors, which indicates that increased MTNR1B mRNA expression is not a general trait of patients with T2DM167.

Melatonin target tissues beyond β-cells

Many studies focused their attention on the pancreas as the main melatonin target tissue involved in glucose homeostasis; however, the involvement of several other peripheral tissues, and in particular the brain, should also be considered (Fig. 1). In this context, it is of interest to consider the increased MTNR1B mRNA expression observed in rs10830963 risk allele carriers, an effect that was proposed to be specific for pancreatic β-cells3. Molecular studies provided the first mechanistic hypothesis by proposing that the region surrounding the rs10830963 SNP is targeted by the forkhead box protein A2 (FOXA2) transcription factor and exhibits enhancer activity168. This enhancer activity was proposed to be increased specifically in human islets by the binding of the transcription factor neurogenic differentiation 1 (NeuroD1) to the rs10830963 sequence of risk allele carriers. Although of interest, these data are not conclusive in their current state. First, it is difficult to demonstrate β-cell specificity as NeuroD1 is also expressed in the brain and digestive tract, two tissues known to express MT2 receptors169. Second, the specificity of the enhancer effect for the transcription of the MTNR1B gene is not demonstrated, and enhancers are well-known to target loci located hundreds of kb away within large-scale topologically associating domains, implying that the transcription of other genes could be modulated by this enhancer.

The idea of modified gene expression of other genes in rs10830963 risk allele carriers is supported by the observed modification of the melatonin secretion profile in risk allele carriers82. This interesting observation reinforces the potential importance of central effects of melatonin in relation to glucose homeostasis regulation, as melatonin is principally secreted by the pineal gland (Fig. 1). As melatonin synthesis is under the control of the biological master clock in the SCN, this finding suggests that the central circadian clock might be modified in risk allele carriers. Importantly, parameters related to the sleep–wake cycle seem to be unaltered in carriers of the risk allele82. Taken together, the nature of the most relevant melatonin target tissue to explain the effects of melatonin on glucose homeostasis in humans remains unknown based on currently available data. In light of the widespread modulatory role of melatonin, the involvement of central and peripheral tissues would not be surprising.

Timing in the melatonergic system

The importance of appropriate timing is another distinctive feature of the melatonergic system, which is reflected by the delayed, circadian and seasonal effects of melatonin. This notion is first of all based on the circadian secretion profile of melatonin, with high levels being secreted during the night-time. It is important to point out that sampling of melatonin in human plasma is not always compatible with routine blood withdrawal practice, which typically occurs in the morning between 08:00 h and 11:00 h, when melatonin levels are low and have no predictive value for night-time levels. Possible alternatives are the measurement of melatonin in saliva during the night or its metabolite 6-sulfatoxy-melatonin in the first morning urine, which allows a good estimation of the melatonin levels during the previous night. The importance of timing is not only true for the production of melatonin but also for the sensitivity of the body to melatonin, which is not constant over the 24-hour cycle170,171. The lengthened melatonin secretion profile in the morning in rs10830963 risk allele carriers is interesting in this respect82. Based on the nocturnal melatonin secretion profile, the action of melatonin (that is, regulation of glucose homeostasis) would be expected to occur in the same time frame — at night-time. However, melatonin is also known to have delayed effects, as illustrated by the sensitization of the cAMP system following the long-term presence of melatonin152 or the modulation of GLUT4 expression upon long-term melatonin treatment6,57.

Two studies published in 2018 further support the notion that the effect of melatonin receptor activation might also be visible during the daytime, at a time when the regulation of plasma concentrations of insulin is most important in humans. First, nocturnal activation of the MT1 receptor was shown to modulate insulin sensitivity during the day in mice via the regulation of the transcription of PI3K55. Second, the association between loss of spontaneous MT2 receptor activity with increased T2DM risk, as shown for carriers of rare MT2 mutations, is not dependent on the presence of melatonin80. These observations change our understanding of the time windows at which nocturnal melatonin might modulate insulin sensitivity, which warrants further attention. In conclusion, the notion of timing is an important element of the melatonergic system that has to be considered and translated into clinical practice.

MT1 in dysregulated glucose homeostasis

Since the discovery of the association of common and rare SNPs in the MTNR1B locus with the risk of T2DM, research efforts have focused on the MT2 receptor as the primary melatonin target relevant for glucose homeostasis. In turn, only minor attention has been given to the potential role of the MT1 receptor. Studies in rodent cells and rodent models suggest that MT1 and MT2 receptors are at least equally important and an extensive characterization of MT1-knockout mice that was published in 2017 provided supportive evidence for the importance of this receptor in the regulation of insulin sensitivity55.

Expression of MT1 is more abundant than that of MT2 and both have similar expression patterns at central and peripheral sites31. A notable difference between MT1 and MT2 receptors is their regulation. Whereas MT2 has been proposed to desensitize the cAMP response upon melatonin stimulation172, activation of MT1 has been associated with super-sensitization of the cAMP response during the subsequent period of withdrawal173. The sensitization of the cAMP system is probably involved in the stimulatory effect of melatonin on insulin secretion upon long-term treatment of INS-1 cells and human pancreatic islets with melatonin53,62. Intriguingly, no association has been reported so far between the MTNR1A gene (which encodes the MT1 receptor) and T2DM or obesity. The associations of rare MTNR1A variants with these diseases and traits are currently unknown. Overlapping expression patterns, together with redundant signalling functions, is intriguing in light of the apparently different contributions of both receptors in the regulation of glucose homeostasis in mice174. Apart from differences in receptor regulation, different subcellular localization of MT1 and MT2 might be another possible explanation, which has emerged following the demonstration of the mitochondrial localization of MT1 receptors21,175. Furthermore, differences in the molecular microenvironment, such as differences in the capacity of heterodimer formation between MT1 and MT2 (refs176,177), should be also considered, in our opinion.

Taken together, research on melatonin and glucose homeostasis made considerable progress in the past 5 years. To resolve the contradictory findings outlined here, a broader vision in terms of potential signalling pathways, physiological actions and target tissues, together with a better integration of the rhythmic-behaviour of the melatonergic system, has to be considered. The latest results in this area provide some interesting new lines of investigation to be followed.

Strategies to manage risk of T2DM

GPCRs are popular drug targets, as ~30% of the currently marketed drugs act on GPCRs178,179. Currently marketed drugs targeting melatonin receptors are indicated for insomnia (ramelteon), insomnia in the elderly (slow-release melatonin preparation), non-24-h sleep-wake disorder in totally blind individuals (tasimelteon) and major depressive disorders (agomelatine)180. All these compounds are nonselective agonists of the MT1 and MT2 receptors. Agomelatine belongs to a new class of melatonin receptor ligands as it displays two complementary activities by targeting not only melatonin receptors but also serotonin 5-HT2C receptors181. Other multi-target-directed ligands are currently under evaluation. For instance, piromelatine (a melatonin receptor and 5-HT1A and 5-HT1D receptor agonist) has already completed a phase II clinical trial for insomnia and is in a phase IIb clinical trial to treat mild Alzheimer disease182,183. Metabolic diseases, including T2DM, are currently not among the clinical indications for melatonin receptor ligands. In the course of controlled clinical trials designed for other indications (such as insomnia and depression), altered glucose homeostasis, insulin secretion or T2DM risk were not reported as measured secondary outcomes. Several small-scale controlled clinical studies have been carried out and will be discussed in the following paragraphs.

Apart from medication, lifestyle recommendations are an essential component of the repertoire of therapeutic options to combat the epidemics of obesity and T2DM. Proper adjustment of the individual circadian rhythm to the environment is part of these recommendations and the hormone melatonin, as a reliable output and input signal of the central circadian system, is an obvious parameter to be considered. Personalization of lifestyle recommendations is becoming more and more possible as a result of the increasing number of genetic variants that have been associated with altered metabolic traits and T2DM risk. Based on the gene variants present in a patient, specific recommendations can be made. A confounding parameter to be taken into account when providing therapeutic recommendations, and also diagnosis, is the notable rate of melatonin self-medication, as the number of countries allowing commercialization of melatonin without any medical prescription is quickly increasing, including the USA and several European countries184,185.

Pharmacological targeting

The availability of melatonin depends on the country. In several countries, such as the USA, melatonin is considered a dietary supplement and can be purchased in drug stores without any prescription. In other countries, such as the UK, melatonin can only be obtained with a prescription. Melatonin is popular for the promotion of improved sleep initiation and for fast adjustment in situations of circadian misalignment (such as jet-lag when travelling over several time zones), but also as a prophylactic anti-ageing treatment and as a preventive treatment for neurodegenerative diseases and cancer58. The rationale behind melatonin supplementation is based on the observation that the amplitude and thus the strength of the natural melatonin rhythm tends to decline with age58. This decline is even further accelerated in people with neurodegenerative diseases such as Alzheimer disease, Parkinson disease and Huntington disease1. An estimated 3.1 million adults in the USA (1.3% of US adults) use exogenous melatonin184, 1.4 million packages are sold in France per year186 and the average yearly cost of melatonin prescription for insomnia and anxiety is more than £22 million in the UK187. These facts, in conjunction with the increased numbers of night shift workers (in the UK, the number of people who work night shifts increased by 275,000 (9%) between 2011 and 2016)188 and the high numbers of individuals with late-night eating behaviour syndrome (the prevalence of night-eating syndrome is estimated to be 1.1–1.5% in the general population189) means that the understanding of human melatonin receptor function and its role in glucose homeostasis is of primary importance for public health care.

Should treatment with melatonin or melatonin receptor ligands be considered in patients with T2DM? Solely based on the finding that patients with T2DM have lower night-time melatonin levels than participants without T2DM93,95, melatonin replacement could be advised. As detailed in previous sections, preclinical data on the use of melatonin in these patients are conflicting, ranging from improved glycaemic control to impaired glucose tolerance depending on the protocol applied (for example, chronic versus acute administration)190. Clearly, further studies are necessary before considering melatonin administration in patients with T2DM. In this context, it is worth mentioning that not only is the right drug dosage important for melatonin but also the time of administration and the duration of the effect. Melatonin is generally taken once a day, 1–2 hours before an individual’s bedtime to mimic, at least partially, the natural night-time melatonin peak. As the amplitude of the melatonin rhythm varies between different individuals, and given that this parameter is rarely determined before starting melatonin self-medication, achievement of supra-physiological levels has to be assumed in many cases82. This high level of uncontrolled melatonin supplementation has to be taken into account when evaluating the individual risk of developing T2DM in those who are obese and are at high risk of developing T2DM.

Another aspect relevant to pharmacological intervention in patients with T2DM who are carriers of the rs10830963 risk allele might come from their increased response to approved T2DM drugs compared with non-carriers191. An increased insulin response to glucagon-like peptide 1 (GLP1) has been reported in carriers of the rs10830963 risk allele, which suggests that these patients might benefit from treatment with GLP1 agonists191. Furthermore, the deleterious effect of the rs10830963 SNP on β-cell function seems to not only persist over time but to worsen with time in individuals with impaired glucose tolerance3,192, which underscores the benefits of intervening pharmacologically at an early stage, before glycaemic deterioration occurs.

Altogether, there is currently not enough clinical evidence to consider prescribing melatonin to patients with T2DM in general. The identification of frequent variants within the MTNR1B locus has opened the possibility of personalized medication, as will be discussed in the next section. Independent of the genetic background, the high level of self-medication, with several million people taking melatonin on a daily basis, has to be surveyed closely by health authorities to minimize the potential risks until more consistent clinical data on melatonin treatment in humans are available.

Personalized health care

Classification of the population according to their genotype is of high clinical and public health relevance as health-care recommendations for associated risks (for example, T2DM and GDM) can be refined in a genotype-specific manner. Among the multiple variants identified at the MTNR1B locus, the rs10830963 variant is of particular interest because of its high prevalence (MAF ~30%) and robust association with T2DM risk3. On the basis of the assumption that increased MT2 receptor function is causally linked to the increased T2DM risk seen in rs10830963 risk allele carriers, it would be expected that melatonin treatment would worsen the T2DM phenotype. This effect has been observed upon acute melatonin treatment in the morning96, and both morning and evening hours97, but not upon long-term melatonin administration at bedtime98,100, which improved glycaemic control. However, the status of the rs10830963 locus was not defined in these studies. When separating rs10830963 risk from non-risk carriers, acute morning melatonin administration worsened glucose tolerance in risk allele carriers, while no difference on glucose tolerance was observed in the evening between the two genotypes193. Longer administration of evening melatonin over a 3-month period resulted in decreased first-phase insulin release and an increase in glucose concentrations in both risk and non-risk carriers, with more pronounced affects in the risk carriers56. In regard to the baseline glucose levels in the risk allele carriers, no impairment was observed after 3 months of melatonin administration and hence further insights in the regulation of endogenous melatonin levels might be more informative.

Taken together, rs10830963 risk allele carriers seem to be particularly glucose intolerant in the morning when their endogenous melatonin levels are still elevated, but their risk of glucose intolerance in the evening is similar to that of carriers of non-risk alleles. Consistently, the increased risk of T2DM for the risk allele carriers was more pronounced in early risers than in late risers, further indicating that the decreased glucose tolerance in the morning might be casually linked to the elevated endogenous melatonin levels82. Another study focused on the evening hours by evaluating the influence of the rs10830963 risk allele on glucose tolerance in the context of an early versus a late dinner with the idea that the late dinner falls under conditions when melatonin levels start to rise. On the basis of the assumption that elevated melatonin levels are detrimental for glucose tolerance, glucose tolerance should be worse at late dinners. This effect was observed in the general population and in homozygous rs10830963 risk allele carriers but surprisingly not in homozygous non-risk carriers194. The origin of the difference between the two phenotypes is not clear. However, rs10830963 risk allele carriers should thus avoid late dinners. Similarly, concurrence of meals and melatonin treatment should be avoided in risk allele carriers who take melatonin or synthetic melatonin receptor agonists for medical or prophylactic reasons, which is typically in the evening 1–2 hours before going to bed.

Clinical trials testing the effect of risk alleles on lifestyle interventions might provide further insights. Several intervention studies have evaluated the effect of the rs10830963 SNP on weight loss and improvement of plasma lipid parameters77,78,85,86,87,88. Unfortunately, these studies mainly monitored obesity-related parameters for which no clear association with the rs10830963 SNP exists, as discussed previously. Carriers of the rs10830963 risk allele are particularly susceptible to putative desynchronizing environmental agents, such as noise pollution, as a positive correlation between noise and changes in HbA1c levels over 8 years has been observed195. A possible explanation for this observation is the increased susceptibility of risk carriers for morning glucose intolerance due to their extended melatonin secretion in the morning hours, which reflects the increased risk of T2DM in risk allele carriers who are early risers82.

Unfortunately, controlled studies of clinical melatonin treatment and interventional studies are not available for carriers of rare MTNR1B variants with a loss-of-function phenotype (which is associated with an increased risk of T2DM). Such studies would be extremely informative to solve the controversy of the association of loss-of function or gain-of-function hypotheses with T2DM risk but are limited by the low prevalence (MAF <0.1%)75 of these very rare variants.

In conclusion, many insights have been obtained from melatonin treatment studies of rs10830963 risk allele carriers, who seem to be particularly glucose intolerant in the morning when their endogenous levels are still elevated. Refining these studies in terms of treatment protocols in addition to clinical trials to clarify the effect of risk alleles on lifestyle interventions will certainly provide further valuable insights in the future.

Perspectives

The characterization of the phenotype of carriers of the rs10830963 risk allele turned out to be particularly fruitful; however, several important questions related to this SNP remain to be solved. In rs10830963 risk allele carriers, no statistically significant association was reported with several circadian traits, except for a reported association with changes in the melatonin secretion profile82. As melatonin is the most reliable output signal of the central circadian rhythm-generating clock, the observed changes in offset and duration of melatonin synthesis are probably due to modifications in the central circadian machinery. Future studies will have to clarify this point. An important question to be addressed is whether normalizing the melatonin secretion profile will also normalize glucose metabolism of the risk allele carriers. Notably, several therapeutic options are available to test this hypothesis. For example, as melatonin synthesis is inhibited by light, a light therapy early in the morning can be applied under controlled laboratory conditions. Furthermore, β-blockers could be used early in the morning to block β-adrenergic receptors that regulate melatonin synthesis in the pineal gland196. Alternatively, the effect of melatonin could be inhibited by treating risk allele carriers with the melatonin receptor-specific antagonist S20928 (ref.197). Delaying the time of breakfast for risk allele carriers might be another measure to decrease their risk of T2DM, which can be tested easily.

Further genetic studies are also likely to provide valuable insights in the future, in particular, in defining the effect of the other players in the melatonin system, notably the two genes involved in the melatonin synthesis pathway (AANAT and ASMT) and MTNR1A. Loss-of-function mutants have been detected for ASMT in people with autism spectrum disorders198,199. Unfortunately, their metabolic phenotype has not been reported. A similar situation exists for MT1, for which several loss-of-function mutations have been identified, but no data on the metabolic phenotype is available200. For the MTNR1A gene, copy number variants might also be relevant as a bioinformatic study revealed that this gene is ranked fourth among >100 clinically relevant genes encoding GPCRs in terms of the number of observed gene duplications187. Given that MT1 and MT2 receptors show overlapping expression profiles, often with higher expression levels for MT1, together with the high redundancy between these two receptors in terms of cellular functions, the absence of frequent gene variants for MT1 that are associated with the risk of T2DM remains a mystery that needs to be solved to better understand the distinctive features of the MT2 receptor.

In light of the evidence presented in this Review, the use of melatonin both as a supplement and as a medicine needs to be carefully re-evaluated as it might significantly affect glucose homeostasis. Improved selectivity of future medication (for example, by signalling pathway-biased ligands) might be of interest in this context. The identification of an association of loss-of-function of specific melatonin signalling pathways with T2DM risk80 and the description of the first biased ligands for melatonin receptors are promising steps in this direction175. Combination therapies are also likely to gain more importance in the future, an idea that fits well with the notion that melatonin is a ‘modulator’ rather than a ‘driver’ for many physiological functions91,99. The capacity of melatonin receptors to form oligomeric complexes composed of MT1 and MT2 receptors or between melatonin receptors and other GPCRs, such as the 5-HT2C receptor, warrants more attention in the future as it opens new options for multi-target-directed ligands.

Finally, along with the ‘new’ receptor-dependent functions of melatonin, the ‘old’ function as an antioxidant should not be ignored, as this is a function that emerged a long time before melatonin receptors appeared during evolution7. Several reports suggest that melatonin might have important protective receptor-dependent and receptor-independent effects on mitochondrial functions159,201 that could be relevant for the metabolic effects of melatonin, an aspect that should be studied in more detail in the future.

Conclusions

Since the discovery of the association of frequent SNPs within the MTNR1B locus with the risk of T2DM in 2009, much progress has been made in the understanding of the interplay between melatonin and glucose homeostasis in humans. The effect of the risk allele rs10830963 SNP probably starts early during the development of prediabetic fasting hyperglycaemia by affecting insulin secretion. Obesity, an important risk factor for T2DM development, seems to not be associated with the rs10830963 SNP in adults but might have a role in fetal birth weight through effects on maternal melatonin secretion.

Controlled clinical studies, particularly those including carriers of the rs10830963 SNP, provided important information on the effect of melatonin treatment on glucose homeostasis and started to unravel the underlying molecular mechanisms of the association, which includes increased MTNR1B mRNA levels, a modified melatonin secretion profile and possibly other effects due to the enhancer activity of the region surrounding the rs10830963 SNP. The relative contribution of these different effects to the association of this SNP with T2DM remains to be established. Importantly, personalized lifestyle recommendations are starting to emerge based on interventional studies. Rare mutations in the MTNR1B gene provided further important insights, revealing that loss-of-function, and not gain-of-function of particular signalling functions of the MT2 receptor is associated with the risk of T2DM. These in vitro data will have to be validated in patient tissues and in vivo. Studies of defects in a subset of MT2 signalling functions will eventually open the avenues for pathway-specific personalized therapeutic interventions.

Based on the existing data, it is yet to be established whether melatonin treatment for patients with T2DM causes beneficial or adverse effects. However, the high level of self-medication, with several million people taking melatonin on a daily basis, warrants a rapid answer to this question. To fully consolidate the role of the melatonergic system in glucose homeostasis in humans, the integration of the unique features of the melatonergic system with a broader vision in terms of the other components of the melatonergic system could result in opportunities for new therapeutic approaches in the treatment of T2DM, including multi-target-directed ligands.

References

  1. 1.

    Jockers, R. et al. Update on melatonin receptors: IUPHAR review 20. Br. J. Pharmacol. 173, 2702–2725 (2016).

  2. 2.

    Bouatia-Naji, N. et al. A variant near MTNR1B is associated with increased fasting plasma glucose levels and type 2 diabetes risk. Nat. Genet. 41, 89–94 (2009).

  3. 3.

    Lyssenko, V. et al. Common variant in MTNR1B associated with increased risk of type 2 diabetes and impaired early insulin secretion. Nat. Genet. 41, 82–88 (2009).

  4. 4.

    Prokopenko, I. et al. Variants in MTNR1B influence fasting glucose levels. Nat. Genet. 41, 77–81 (2009).

  5. 5.

    Zisapel, N. New perspectives on the role of melatonin in human sleep, circadian rhythms and their regulation. Br. J. Pharmacol. 175, 3190–3199 (2018).

  6. 6.

    Cipolla-Neto, J., Amaral, F. G., Afeche, S. C., Tan, D. X. & Reiter, R. J. Melatonin, energy metabolism, and obesity: a review. J. Pineal Res. 56, 371–381 (2014).

  7. 7.

    Manchester, L. C. et al. Melatonin: an ancient molecule that makes oxygen metabolically tolerable. J. Pineal Res. 59, 403–419 (2015).

  8. 8.

    Bernard, M. et al. Melatonin synthesis pathway: circadian regulation of the genes encoding the key enzymes in the chicken pineal gland and retina. Reprod. Nutr. Dev. 39, 325–334 (1999).

  9. 9.

    Arendt, J. Melatonin and the pineal gland: influence on mammalian seasonal and circadian physiology. Rev. Reprod. 3, 13–22 (1998).

  10. 10.

    Tricoire, H., Locatelli, A., Chemineau, P. & Malpaux, B. Melatonin enters the cerebrospinal fluid through the pineal recess. Endocrinology 143, 84–90 (2002).

  11. 11.

    Legros, C., Chesneau, D., Boutin, J. A., Barc, C. & Malpaux, B. Melatonin from cerebrospinal fluid but not from blood reaches sheep cerebral tissues under physiological conditions. J. Neuroendocrinol. 26, 151–163 (2014).

  12. 12.

    Pevet, P. & Challet, E. Melatonin: both master clock output and internal time-giver in the circadian clocks network. J. Physiol. Paris 105, 170–182 (2011).

  13. 13.

    Pevet, P., Klosen, P. & Felder-Schmittbuhl, M. P. The hormone melatonin: animal studies. Best Pract. Res. Clin. Endocrinol. Metab. 31, 547–559 (2017).

  14. 14.

    Zlotos, D. P., Jockers, R., Cecon, E., Rivara, S. & Witt-Enderby, P. A. MT1 and MT2 melatonin receptors: ligands, models, oligomers, and therapeutic potential. J. Med. Chem. 57, 3161–3185 (2014).

  15. 15.

    Baba, K. et al. Heteromeric MT1/MT2 melatonin receptors modulate photoreceptor function. Sci. Signal. 6, ra89 (2013).

  16. 16.

    Conti, A. et al. Evidence for melatonin synthesis in mouse and human bone marrow cells. J. Pineal Res. 28, 193–202 (2000).

  17. 17.

    Markus, R. P., Fernandes, P. A., Kinker, G. S., da Silveira Cruz-Machado, S. & Marcola, M. Immune-pineal axis — acute inflammatory responses coordinate melatonin synthesis by pinealocytes and phagocytes. Br. J. Pharmacol. 175, 3239–3250 (2017).

  18. 18.

    Yi, W. J. & Kim, T. S. Melatonin protects mice against stress-induced inflammation through enhancement of M2 macrophage polarization. Int. Immunopharmacol. 48, 146–158 (2017).

  19. 19.

    Bertrand, P. P., Polglaze, K. E., Bertrand, R. L., Sandow, S. L. & Pozo, M. J. Detection of melatonin production from the intestinal epithelium using electrochemical methods. Curr. Pharm. Des. 20, 4802–4806 (2014).

  20. 20.

    Chen, C. Q., Fichna, J., Bashashati, M., Li, Y. Y. & Storr, M. Distribution, function and physiological role of melatonin in the lower gut. World J. Gastroenterol. 17, 3888–3898 (2011).

  21. 21.

    Suofu, Y. et al. Dual role of mitochondria in producing melatonin and driving GPCR signaling to block cytochrome c release. Proc. Natl Acad. Sci. USA 114, E7997–E8006 (2017).

  22. 22.

    He, C. et al. Mitochondria synthesize melatonin to ameliorate its function and improve mice oocyte’s quality under in vitro conditions. Int. J. Mol. Sci. 17, 939 (2016).

  23. 23.

    Tan, D. X. et al. Mitochondria and chloroplasts as the original sites of melatonin synthesis: a hypothesis related to melatonin’s primary function and evolution in eukaryotes. J. Pineal Res. 54, 127–138 (2013).

  24. 24.

    Dubocovich, M. L. et al. International union of basic and clinical pharmacology. LXXV. Nomenclature, classification, and pharmacology of G protein-coupled melatonin receptors. Pharmacol. Rev. 62, 343–380 (2010).

  25. 25.

    Bondi, C. D. et al. MT1 melatonin receptor internalization underlies melatonin-induced morphologic changes in Chinese hamster ovary cells and these processes are dependent on Gi proteins, MEK 1/2 and microtubule modulation. J. Pineal Res. 44, 288–298 (2008).

  26. 26.

    Hong, L. J. et al. Valproic acid influences MTNR1A intracellular trafficking and signaling in a beta-arrestin 2-dependent manner. Mol. Neurobiol. 53, 1237–1246 (2016).

  27. 27.

    Levoye, A. et al. The orphan GPR50 receptor specifically inhibits MT(1) melatonin receptor function through heterodimerization. EMBO J. 25, 3012–3023 (2006).

  28. 28.

    Benleulmi-Chaachoua, A. et al. Protein interactome mining defines melatonin MT1 receptors as integral component of presynaptic protein complexes of neurons. J. Pineal Res. 60, 95–108 (2016).

  29. 29.

    Guillaume, J. L. et al. The PDZ protein mupp1 promotes Gi coupling and signaling of the Mt1 melatonin receptor. J. Biol. Chem. 283, 16762–16771 (2008).

  30. 30.

    Maurice, P. et al. Molecular organization and dynamics of the melatonin MT receptor/RGS20/G(i) protein complex reveal asymmetry of receptor dimers for RGS and G(i) coupling. EMBO J. 29, 3646–3659 (2010).

  31. 31.

    Cecon, E., Oishi, A. & Jockers, R. Melatonin receptors: molecular pharmacology and signalling in the context of system bias. Br. J. Pharmacol. 175, 3263–3280 (2017).

  32. 32.

    Oishi, A., Cecon, E. & Jockers, R. Melatonin receptor signaling: impact of receptor oligomerization on receptor function. Int. Rev. Cell. Mol. Biol. 338, 59–77 (2018).

  33. 33.

    Chen, L. et al. Melatonin receptor type 1 signals to extracellular signal-regulated kinase 1 and 2 via Gi and Gs dually coupled pathways in HEK-293 cells. Biochemistry 53, 2827–2839 (2014).

  34. 34.

    Chan, A. S. et al. Melatonin mt1 and MT2 receptors stimulate c-Jun N-terminal kinase via pertussis toxin-sensitive and -insensitive G proteins. Cell. Signal. 14, 249–257 (2002).

  35. 35.

    Sack, R. L., Brandes, R. W., Kendall, A. R. & Lewy, A. J. Entrainment of free-running circadian rhythms by melatonin in blind people. N. Engl. J. Med. 343, 1070–1077 (2000).

  36. 36.

    Isherwood, C. M., Van der Veen, D. R., Johnston, J. D. & Skene, D. J. Twenty-four-hour rhythmicity of circulating metabolites: effect of body mass and type 2 diabetes. FASEB J. 31, 5557–5567 (2017).

  37. 37.

    Qian, J. & Scheer, F. Circadian system and glucose metabolism: implications for physiology and disease. Trends Endocrinol. Metab. 27, 282–293 (2016).

  38. 38.

    Bass, J. in A Time for Metabolism and Hormones (eds Sassone-Corsi, P. & Christen, Y.) 25–32 (Springer International Publishing, 2016).

  39. 39.

    von Gall, C., Weaver, D. R., Kock, M., Korf, H. W. & Stehle, J. H. Melatonin limits transcriptional impact of phosphoCREB in the mouse SCN via the Mel1a receptor. Neuroreport 11, 1803–1807 (2000).

  40. 40.

    Mc Arthur, A., Hunt, A. & Gillette, M. Melatonin action and signal transduction in the rat suprachiasmatic circadian clock: activation of protein kinase C at dusk and dawn. Endocrinology 138, 627–634 (1997).

  41. 41.

    Hunt, A. E., AlGhoul, W. M., Gillette, M. U. & Dubocovich, M. L. Activation of MT2 melatonin receptors in rat suprachiasmatic nucleus phase advances the circadian clock. Amer. J. Physiol. Cell Physiol. 280, C110–C118 (2001).

  42. 42.

    Nelson, C. S., Marino, J. L. & Allen, C. N. Melatonin receptors activate heteromeric G-protein coupled Kir3 channels. Neuroreport 7, 717–720 (1996).

  43. 43.

    Hablitz, L. M. et al. GIRK channels mediate the nonphotic effects of exogenous melatonin. J. Neurosci. 35, 14957–14965 (2015).

  44. 44.

    Pfeffer, M., Rauch, A., Korf, H. W. & von Gall, C. The endogenous melatonin (MT) signal facilitates reentrainment of the circadian system to light-induced phase advances by acting upon MT2 receptors. Chronobiol. Int. 29, 415–429 (2012).

  45. 45.

    Nagy, A. D. et al. Melatonin adjusts the expression pattern of clock genes in the suprachiasmatic nucleus and induces antidepressant-like effect in a mouse model of seasonal affective disorder. Chronobiol. Int. 32, 447–457 (2015).

  46. 46.

    Kandalepas, P. C., Mitchell, J. W. & Gillette, M. U. Melatonin signal transduction pathways require E-box-mediated transcription of Per1 and Per2 to reset the SCN clock at dusk. PLOS ONE 11, e0157824 (2016).

  47. 47.

    Ha, E. et al. Melatonin stimulates glucose transport via insulin receptor substrate-1/phosphatidylinositol 3-kinase pathway in C2C12 murine skeletal muscle cells. J. Pineal Res. 41, 67–72 (2006).

  48. 48.

    Poon, A. M., Choy, E. H. & Pang, S. F. Modulation of blood glucose by melatonin: a direct action on melatonin receptors in mouse hepatocytes. Biol. Signals Recept. 10, 367–379 (2001).

  49. 49.

    Sauer, L. A., Dauchy, R. T. & Blask, D. E. Melatonin inhibits fatty acid transport in inguinal fat pads of hepatoma 7288CTC-bearing and normal Buffalo rats via receptor-mediated signal transduction. Life Sci. 68, 2835–2844 (2001).

  50. 50.

    Muhlbauer, E., Albrecht, E., Bazwinsky-Wutschke, I. & Peschke, E. Melatonin influences insulin secretion primarily via MT(1) receptors in rat insulinoma cells (INS-1) and mouse pancreatic islets. J. Pineal Res. 52, 446–459 (2012).

  51. 51.

    Zalatan, F., Krause, J. A. & Blask, D. E. Inhibition of isoproterenol-induced lipolysis in rat inguinal adipocytes in vitro by physiological melatonin via a receptor-mediated mechanism. Endocrinology 142, 3783–3790 (2001).

  52. 52.

    Stumpf, I., Muhlbauer, E. & Peschke, E. Involvement of the cGMP pathway in mediating the insulin-inhibitory effect of melatonin in pancreatic beta-cells. J. Pineal Res. 45, 318–327 (2008).

  53. 53.

    Kemp, D. M., Ubeda, M. & Habener, J. F. Identification and functional characterization of inelatonin Mel 1a receptors in pancreatic beta cells: potential role in incretin-mediated cell function by sensitization of cAMP signaling. Mol. Cell. Endocrinol. 191, 157–166 (2002).

  54. 54.

    Shieh, J. M., Wu, H. T., Cheng, K. C. & Cheng, J. T. Melatonin ameliorates high fat diet-induced diabetes and stimulates glycogen synthesis via a PKCzeta-Akt-GSK3beta pathway in hepatic cells. J. Pineal Res. 47, 339–344 (2009).

  55. 55.

    Owino, S. et al. Nocturnal activation of melatonin receptor type 1 signaling modulates diurnal insulin sensitivity via regulation of PI3K activity. J. Pineal Res. 64, e12462 (2018).

  56. 56.

    Tuomi, T. et al. Increased melatonin signaling is a risk factor for type 2 diabetes. Cell Metab. 23, 1067–1077 (2016).

  57. 57.

    Brydon, L., Petit, L., Delagrange, P., Strosberg, A. D. & Jockers, R. Functional expression ofmt2 (mel1b) melatonin receptors in human paz6 adipocytes. Endocrinology 142, 4264–4271 (2001).

  58. 58.

    Karamitri, A., Renault, N., Clement, N., Guillaume, J. L. & Jockers, R. Minireview: toward the establishment of a link between melatonin and glucose homeostasis: association of melatonin MT2 receptor variants with type 2 diabetes. Mol. Endocrinol. 27, 1217–1233 (2013).

  59. 59.

    Peschke, E., Bahr, I. & Muhlbauer, E. Melatonin and pancreatic islets: interrelationships between melatonin, insulin and glucagon. Int. J. Mol. Sci. 14, 6981–7015 (2013).

  60. 60.

    Bahr, I., Muhlbauer, E., Schucht, H. & Peschke, E. Melatonin stimulates glucagon secretion in vitro and in vivo. J. Pineal Res. 50, 336–344 (2011).

  61. 61.

    Ramracheya, R. D. et al. Function and expression of melatonin receptors on human pancreatic islets. J. Pineal Res. 44, 273–279 (2008).

  62. 62.

    Costes, S., Boss, M., Thomas, A. P. & Matveyenko, A. V. Activation of melatonin signaling promotes beta-cell survival and function. Mol. Endocrinol. 29, 682–692 (2015).

  63. 63.

    Ruiz, L. et al. Proteasomal degradation of the histone acetyl transferase p300 contributes to beta-cell injury in a diabetes environment. Cell Death Dis. 9, 600 (2018).

  64. 64.

    Zibolka, J., Muhlbauer, E. & Peschke, E. Melatonin influences somatostatin secretion from human pancreatic delta-cells via MT1 and MT2 receptors. J. Pineal Res. 58, 198–209 (2015).

  65. 65.

    Zibolka, J., Bazwinsky-Wutschke, I., Muhlbauer, E. & Peschke, E. Distribution and density of melatonin receptors in human main pancreatic islet cell types. J. Pineal Res. 65, e12480 (2018).

  66. 66.

    Bartness, T. J., Powers, J. B., Hastings, M. H., Bittman, E. L. & Goldman, B. D. The timed infusion paradigm for melatonin delivery: what has it taught us about the melatonin signal, its reception, and the photoperiodic control of seasonal responses? J. Pineal Res. 15, 161–190 (1993).

  67. 67.

    Reiter, R. J. Photoperiod: its importance as an impeller of pineal and seasonal reproductive rhythms. Int. J. Biometeorol. 24, 57–63 (1980).

  68. 68.

    Peschke, E. & Muhlbauer, E. New evidence for a role of melatonin in glucose regulation. Best Pract. Res. Clin. Endocrinol. Metab. 24, 829–841 (2010).

  69. 69.

    Wade, G. N. & Bartness, T. J. Effects of photoperiod and gonadectomy on food intake, body weight, and body composition in Siberian hamsters. Am. J. Physiol. 246, R26–R30 (1984).

  70. 70.

    Le Gouic, S. et al. Characterization of a melatonin binding site in Siberian hamster brown adipose tissue. Eur. J. Pharmacol. 339, 271–278 (1997).

  71. 71.

    Prunet, M. B. et al. Evidence for a direct effect of melatonin on mitochondrial genome expression of Siberian hamster brown adipocytes. J. Pineal Res. 30, 108–115 (2001).

  72. 72.

    Lima, F. B. et al. The regulation of insulin action in isolated adipocytes. Role of the periodicity of food intake, time of day and melatonin. Braz. J. Med. Biol. Res. 27, 995–1000 (1994).

  73. 73.

    Contreras-Alcantara, S., Baba, K. & Tosini, G. Removal of melatonin receptor type 1 induces insulin resistance in the mouse. Obesity 18, 1861–1863 (2010).

  74. 74.

    Zhao, J. et al. Examination of all type 2 diabetes GWAS loci reveals HHEX-IDE as a locus influencing pediatric BMI. Diabetes 59, 751–755 (2010).

  75. 75.

    Bonnefond, A. et al. Rare MTNR1B variants impairing melatonin receptor 1B function contribute to type 2 diabetes. Nat. Genet. 44, 297–301 (2012).

  76. 76.

    Yang, J. et al. Genetic association study with metabolic syndrome and metabolic-related traits in a cross-sectional sample and a 10-year longitudinal sample of chinese elderly population. PLOS ONE 9, e100548 (2014).

  77. 77.

    Goni, L. et al. Macronutrient-specific effect of the MTNR1B genotype on lipid levels in response to 2 year weight-loss diets. J. Lipid Res. 59, 155–161 (2018).

  78. 78.

    Goni, L. et al. A circadian rhythm-related MTNR1B genetic variant modulates the effect of weight-loss diets on changes in adiposity and body composition: the POUNDS LOST trial. Eur. J. Nutr. https://doi.org/10.1007/s00394-018-1660-y (2018).

  79. 79.

    Andersson, E. A. et al. MTNR1B G24E variant associates With BMI and fasting plasma glucose in the general population in studies of 22,142 Europeans. Diabetes 59, 1539–1548 (2010).

  80. 80.

    Karamitri, A. et al. Type 2 diabetes-associated variants of the MT2 melatonin receptor affect distinct modes of signaling. Sci. Signal. 11, eaan6622 (2018).

  81. 81.

    Beaumont, R. N. et al. Genome-wide association study of offspring birth weight in 86 577 women identifies five novel loci and highlights maternal genetic effects that are independent of fetal genetics. Hum. Mol. Genet. 27, 742–756 (2018).

  82. 82.

    Lane, J. M. et al. Impact of common diabetes risk variant in MTNR1B on sleep, circadian, and melatonin physiology. Diabetes 65, 1741–1751 (2016).

  83. 83.

    Okatani, Y. et al. Maternal-fetal transfer of melatonin in pregnant women near term. J. Pineal Res. 25, 129–134 (1998).

  84. 84.

    Reiter, R. J., Tan, D. X., Korkmaz, A. & Rosales-Corral, S. A. Melatonin and stable circadian rhythms optimize maternal, placental and fetal physiology. Hum. Reprod. Update 20, 293–307 (2014).

  85. 85.

    Peter, I. et al. Association of type 2 diabetes susceptibility loci with one-year weight loss in the look AHEAD clinical trial. Obesity 20, 1675–1682 (2012).

  86. 86.

    Mirzaei, K. et al. Variants in glucose- and circadian rhythm-related genes affect the response of energy expenditure to weight-loss diets: the POUNDS LOST Trial. Am. J. Clin. Nutr. 99, 392–399 (2014).

  87. 87.

    Goni, L., Cuervo, M., Milagro, F. I. & Martinez, J. A. Gene-gene interplay and gene-diet interactions involving the MTNR1B rs10830963 variant with body weight loss. J. Nutrigenet. Nutrigenom. 7, 232–242 (2014).

  88. 88.

    Grotenfelt, N. E. et al. Interaction between rs10830963 polymorphism in MTNR1B and lifestyle intervention on occurrence of gestational diabetes. Diabetologia 59, 1655–1658 (2016).

  89. 89.

    Lima, F. B. et al. Pinealectomy causes glucose intolerance and decreases adipose cell responsiveness to insulin in rats. Am. J. Physiol. 275, E934–E941 (1998).

  90. 90.

    Nogueira, T. C. et al. Absence of melatonin induces night-time hepatic insulin resistance and increased gluconeogenesis due to stimulation of nocturnal unfolded protein response. Endocrinology 152, 1253–1263 (2011).

  91. 91.

    Oliveira, A. C. et al. Combined treatment with melatonin and insulin improves glycemic control, white adipose tissue metabolism and reproductive axis of diabetic male rats. Life Sci. 199, 158–166 (2018).

  92. 92.

    Champney, T. H., Brainard, G. C., Richardson, B. A. & Reiter, R. J. Experimentally-induced diabetes reduces nocturnal pineal melatonin content in the Syrian hamster. Comp. Biochem. Physiol. A 76, 199–201 (1983).

  93. 93.

    Peschke, E. et al. Diabetic Goto Kakizaki rats as well as type 2 diabetic patients show a decreased diurnal serum melatonin level and an increased pancreatic melatonin-receptor status. J. Pineal Res. 40, 135–143 (2006).

  94. 94.

    Sartori, C. et al. Melatonin improves glucose homeostasis and endothelial vascular function in high-fat diet-fed insulin-resistant mice. Endocrinology 150, 5311–5317 (2009).

  95. 95.

    McMullan, C. J., Schernhammer, E. S., Rimm, E. B., Hu, F. B. & Forman, J. P. Melatonin secretion and the incidence of type 2 diabetes. JAMA 309, 1388–1396 (2013).

  96. 96.

    Cagnacci, A. et al. Influence of melatonin administration on glucose tolerance and insulin sensitivity of postmenopausal women. Clin. Endocrinol. 54, 339–346 (2001).

  97. 97.

    Rubio-Sastre, P., Scheer, F. A., Gomez-Abellan, P., Madrid, J. A. & Garaulet, M. Acute melatonin administration in humans impairs glucose tolerance in both the morning and evening. Sleep 37, 1715–1719 (2014).

  98. 98.

    Garfinkel, D. et al. Efficacy and safety of prolonged-release melatonin in insomnia patients with diabetes: a randomized, double-blind, crossover study. Diabetes Metab. Syndr. Obes. 4, 307–313 (2011).

  99. 99.

    Kadhim, H. M. et al. Effects of melatonin and zinc on lipid profile and renal function in type 2 diabetic patients poorly controlled with metformin. J. Pineal Res. 41, 189–193 (2006).

  100. 100.

    Gonciarz, M. et al. Plasma insulin, leptin, adiponectin, resistin, ghrelin, and melatonin in nonalcoholic steatohepatitis patients treated with melatonin. J. Pineal Res. 54, 154–161 (2013).

  101. 101.

    Sparso, T. et al. G-Allele of intronic rs10830963 in MTNR1B confers increased risk of impaired fasting glycemia and type 2 diabetes through an impaired glucose-stimulated insulin release: studies involving 19,605 Europeans. Diabetes 58, 1450–1456 (2009).

  102. 102.

    Voight, B. F. et al. Twelve type 2 diabetes susceptibility loci identified through large-scale association analysis. Nat. Genet. 42, 579–589 (2010).

  103. 103.

    Dietrich, K. et al. Association and evolutionary studies of the melatonin receptor 1B gene (MTNR1B) in the self-contained population of Sorbs from Germany. Diabet. Med. 28, 1373–1380 (2011).

  104. 104.

    Marouli, E. et al. Evaluating the glucose raising effect of established loci via a genetic risk score. PLOS ONE 12, e0186669 (2017).

  105. 105.

    Sabatti, C. et al. Genome-wide association analysis of metabolic traits in a birth cohort from a founder population. Nat. Genet. 41, 35–46 (2009).

  106. 106.

    Ohshige, T. et al. Association of new loci identified in European genome-wide association studies with susceptibility to type 2 diabetes in the Japanese. PLOS ONE 6, e26911 (2011).

  107. 107.

    Fujita, H. et al. Variations with modest effects have an important role in the genetic background of type 2 diabetes and diabetes-related traits. J. Hum. Genet. 57, 776–779 (2012).

  108. 108.

    Ramos, E. et al. Replication of genome-wide association studies (GWAS) loci for fasting plasma glucose in African-Americans. Diabetologia 54, 783–788 (2011).

  109. 109.

    Liu, C. T. et al. Transferability and fine-mapping of glucose and insulin quantitative trait loci across populations: CARe, the Candidate Gene Association Resource. Diabetologia 55, 2970–2984 (2012).

  110. 110.

    Palmer, N. D. et al. Genetic variants associated with quantitative glucose homeostasis traits translate to type 2 diabetes in mexican americans: the GUARDIAN (Genetics Underlying Diabetes in Hispanics) consortium. Diabetes 64, 1853–1866 (2015).

  111. 111.

    Rönn, T. et al. A common variant in MTNR1B, encoding melatonin receptor 1B, is associated with type 2 diabetes and fasting plasma glucose in Han Chinese individuals. Diabetologia 52, 830–833 (2009).

  112. 112.

    Liu, C. et al. MTNR1B rs10830963 is associated with fasting plasma glucose, HbA1C and impaired beta-cell function in Chinese Hans from Shanghai. BMC Med. Genet. 11, 59 (2010).

  113. 113.

    Hu, C. et al. Effects of GCK, GCKR, G6PC2 and MTNR1B variants on glucose metabolism and insulin secretion. PLOS ONE 5, e11761 (2010).

  114. 114.

    Kan, M. Y. et al. Two susceptible diabetogenic variants near/in MTNR1B are associated with fasting plasma glucose in a Han Chinese cohort. Diabet Med. 27, 598–602 (2010).

  115. 115.

    Takeuchi, F. et al. Common variants at the GCK, GCKR, G6PC2-ABCB11 and MTNR1B loci are associated with fasting glucose in two Asian populations. Diabetologia 53, 299–308 (2010).

  116. 116.

    Salman, M. et al. MTNR1B gene polymorphisms and susceptibility to type 2 diabetes: a pilot study in South Indians. Gene 566, 189–193 (2015).

  117. 117.

    Chambers, J. C. et al. Common genetic variation near melatonin receptor MTNR1B contributes to raised plasma glucose and increased risk of type 2 diabetes among Indian Asians and European Caucasians. Diabetes 58, 2703–2708 (2009).

  118. 118.

    Rees, S. D. et al. Effects of 16 genetic variants on fasting glucose and type 2 diabetes in South Asians: ADCY5 and GLIS3 variants may predispose to type 2 diabetes. PLOS ONE 6, e24710 (2011).

  119. 119.

    Kelliny, C. et al. Common genetic determinants of glucose homeostasis in healthy children: the European Youth Heart Study. Diabetes 58, 2939–2945 (2009).

  120. 120.

    Barker, A. et al. Association of genetic loci with glucose levels in childhood and adolescence: a meta-analysis of over 6,000 children. Diabetes 60, 1805–1812 (2011).

  121. 121.

    Song, J. Y. et al. Association of the rs10830963 polymorphism in MTNR1B with fasting glucose levels in Chinese children and adolescents. Obes. Facts 4, 197–203 (2011).

  122. 122.

    Langlois, C. et al. Evaluating the transferability of 15 European-derived fasting plasma glucose SNPs in Mexican children and adolescents. Sci. Rep. 6, 36202 (2016).

  123. 123.

    Holzapfel, C. et al. Association of a MTNR1B gene variant with fasting glucose and HOMA-B in children and adolescents with high BMI-SDS. Eur. J. Endocrinol. 164, 205–212 (2011).

  124. 124.

    Zheng, C. et al. A common variant in the MTNR1b gene is associated with increased risk of impaired fasting glucose (IFG) in youth with obesity. Obesity 23, 1022–1029 (2015).

  125. 125.

    Reinehr, T. et al. Relationship between MTNR1B (melatonin receptor 1B gene) polymorphism rs10830963 and glucose levels in overweight children and adolescents. Pediatr. Diabetes 12, 435–441 (2011).

  126. 126.

    Soranzo, N. et al. Common variants at 10 genomic loci influence hemoglobin A1(C) levels via glycemic and nonglycemic pathways. Diabetes 59, 3229–3239 (2010).

  127. 127.

    Stancáková, A. et al. Association of 18 confirmed susceptibility loci for type 2 diabetes with indices of insulin release, proinsulin conversion, and insulin sensitivity in 5,327 nondiabetic Finnish men. Diabetes 58, 2129–2136 (2009).

  128. 128.

    Langenberg, C. et al. Common genetic variation in the melatonin receptor 1B gene (MTNR1B) is associated with decreased early-phase insulin response. Diabetologia 52, 1537–1542 (2009).

  129. 129.

    ‘t Hart, L. M. et al. Combined risk allele score of eight type 2 diabetes genes is associated with reduced first-phase glucose-stimulated insulin secretion during hyperglycemic clamps. Diabetes 59, 287–292 (2010).

  130. 130.

    Jonsson, A. et al. Effects of common genetic variants associated with type 2 diabetes and glycemic traits on α- and β-cell function and insulin action in humans. Diabetes 62, 2978–2983 (2013).

  131. 131.

    Prokopenko, I. et al. A central role for GRB10 in regulation of islet function in man. PLOS Genet. 10, e1004235 (2014).

  132. 132.

    Wood, A. R. et al. A Genome-wide association study of IVGTT-based measures of first-phase insulin secretion refines the underlying physiology of type 2 diabetes variants. Diabetes 66, 2296–2309 (2017).

  133. 133.

    Walford, G. A. et al. Common genetic variants differentially influence the transition from clinically defined states of fasting glucose metabolism. Diabetologia 55, 331–339 (2012).

  134. 134.

    Vangipurapu, J. et al. Association of indices of liver and adipocyte insulin resistance with 19 confirmed susceptibility loci for type 2 diabetes in 6,733 non-diabetic Finnish men. Diabetologia 54, 563–571 (2011).

  135. 135.

    Ahlqvist, E. et al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol. 6, 361–369 (2018).

  136. 136.

    Kwak, S. H. et al. A genome-wide association study of gestational diabetes mellitus in Korean women. Diabetes 61, 531–541 (2012).

  137. 137.

    Vlassi, M. et al. The rs10830963 variant of melatonin receptor MTNR1B is associated with increased risk for gestational diabetes mellitus in a Greek population. Hormones 11, 70–76 (2012).

  138. 138.

    Huopio, H. et al. Association of risk variants for type 2 diabetes and hyperglycemia with gestational diabetes. Eur. J. Endocrinol. 169, 291–297 (2013).

  139. 139.

    Wang, Y. et al. Association of six single nucleotide polymorphisms with gestational diabetes mellitus in a Chinese population. PLOS ONE 6, e26953 (2011).

  140. 140.

    Rosta, K. et al. Association study with 77 SNPs confirms the robust role for the rs10830963/G of MTNR1B variant and identifies two novel associations in gestational diabetes mellitus development. PLOS ONE 12, e0169781 (2017).

  141. 141.

    Junior, J. P. et al. The MTNR1B gene polymorphism rs10830963 is associated with gestational diabetes in a Brazilian population. Gene 568, 114–115 (2015).

  142. 142.

    Wu, L., Cui, L., Tam, W. H., Ma, R. C. & Wang, C. C. Genetic variants associated with gestational diabetes mellitus: a meta-analysis and subgroup analysis. Sci. Rep. 6, 30539 (2016).

  143. 143.

    Robitaille, J. & Grant, A. M. The genetics of gestational diabetes mellitus: evidence for relationship with type 2 diabetes mellitus. Genet. Med. 10, 240–250 (2008).

  144. 144.

    Hinton, D. R. et al. Novel localization of a G protein, Gz-alpha, in neurons of brain and retina. J. Neurosci. 10, 2763–2770 (1990).

  145. 145.

    Slominski, R. M., Reiter, R. J., Schlabritz-Loutsevitch, N., Ostrom, R. S. & Slominski, A. T. Melatonin membrane receptors in peripheral tissues: distribution and functions. Mol. Cell Endocrinol. 351, 152–166 (2012).

  146. 146.

    Mulder, H. Melatonin signalling and type 2 diabetes risk: too little, too much or just right? Diabetologia 60, 826–829 (2017).

  147. 147.

    Bonnefond, A., Karamitri, A., Jockers, R. & Froguel, P. The difficult journey from genome-wide association studies to pathophysiology: the melatonin receptor 1B (MT2) paradigm. Cell Metab. 24, 345–347 (2016).

  148. 148.

    Bonnefond, A. & Froguel, P. Disentangling the role of melatonin and its receptor MTNR1B in type 2 diabetes: still a long way to go? Curr. Diab. Rep. 17, 122 (2017).

  149. 149.

    Bonnefond, A. & Froguel, P. The case for too little melatonin signalling in increased diabetes risk. Diabetologia 60, 823–825 (2017).

  150. 150.

    Hardeland, R. Melatonin and the pathologies of weakened or dysregulated circadian oscillators. J. Pineal Res. 62, e12377 (2017).

  151. 151.

    Ben-Dyke, R. Diurnal variation of oral glucose tolerance in volunteers and laboratory animals. Diabetologia 7, 156–159 (1971).

  152. 152.

    Barrett, P., Schuster, C., Mercer, J. & Morgan, P. J. Sensitization: a mechanism for melatonin action in the pars tuberalis. J. Neuroendocrinol. 15, 415–421 (2003).

  153. 153.

    Bach, A. G., Wolgast, S., Muhlbauer, E. & Peschke, E. Melatonin stimulates inositol-1,4,5-trisphosphate and Ca2+ release from INS1 insulinoma cells. J. Pineal Res. 39, 316–323 (2005).

  154. 154.

    Depner, C. M., Melanson, E. L., McHill, A. W. & Wright, K. P. Jr. Mistimed food intake and sleep alters 24-hour time-of-day patterns of the human plasma proteome. Proc. Natl Acad. Sci. USA 115, E5390–E5399 (2018).

  155. 155.

    Simsek, N. et al. Effects of melatonin on islet neogenesis and beta cell apoptosis in streptozotocin-induced diabetic rats: an immunohistochemical study. Domest. Anim. Endocrinol. 43, 47–57 (2012).

  156. 156.

    Kanter, M., Uysal, H., Karaca, T. & Sagmanligil, H. O. Depression of glucose levels and partial restoration of pancreatic beta-cell damage by melatonin in streptozotocin-induced diabetic rats. Arch. Toxicol. 80, 362–369 (2006).

  157. 157.

    de Lima, L. M., dos Reis, L. C. & de Lima, M. A. Influence of the pineal gland on the physiology, morphometry and morphology of pancreatic islets in rats. Braz. J. Biol. 61, 333–340 (2001).

  158. 158.

    Kimple, M. E. et al. Deletion of GalphaZ protein protects against diet-induced glucose intolerance via expansion of beta-cell mass. J. Biol. Chem. 287, 20344–20355 (2012).

  159. 159.

    Suofu, Y., Carlisle, D. L., Vilardaga, J. P. & Friedlander, R. M. Reply to Ahluwalia et al.: Contributions of melatonin receptors are tissue-dependent. Proc. Natl Acad. Sci. USA 115, E1944 (2018).

  160. 160.

    Savaskan, E. et al. Reduced hippocampal MT2 melatonin receptor expression in Alzheimer’s disease. J. Pineal Res. 38, 10–16 (2005).

  161. 161.

    Dubocovich, M. L. Melatonin receptors: role on sleep and circadian rhythm regulation. Sleep Med. 8 (Suppl. 3), 34–42 (2007).

  162. 162.

    Fadista, J. et al. Global genomic and transcriptomic analysis of human pancreatic islets reveals novel genes influencing glucose metabolism. Proc. Natl Acad. Sci. USA 111, 13924–13929 (2014).

  163. 163.

    van de Bunt, M. et al. Transcript expression data from human islets links regulatory signals from genome-wide association studies for type 2 diabetes and glycemic traits to their downstream effectors. PLOS Genet. 11, e1005694 (2015).

  164. 164.

    Segerstolpe, A. et al. Single-cell transcriptome profiling of human pancreatic islets in health and type 2 diabetes. Cell Metab. 24, 593–607 (2016).

  165. 165.

    Thomsen, S. K. et al. Systematic functional characterization of candidate causal genes for type 2 diabetes risk variants. Diabetes 65, 3805–3811 (2016).

  166. 166.

    Gerdin, M. J., Masana, M. I., Ren, D., Miller, R. J. & Dubocovich, M. L. Short-term exposure to melatonin differentially affects the functional sensitivity and trafficking of the hMT(1) and hMT(2) melatonin receptors. J. Pharmacol. Exp. Ther. 304, 931–939 (2003).

  167. 167.

    Solimena, M. et al. Systems biology of the IMIDIA biobank from organ donors and pancreatectomised patients defines a novel transcriptomic signature of islets from individuals with type 2 diabetes. Diabetologia 61, 641–657 (2018).

  168. 168.

    Gaulton, K. J. et al. Genetic fine mapping and genomic annotation defines causal mechanisms at type 2 diabetes susceptibility loci. Nat. Genet. 47, 1415–1425 (2015).

  169. 169.

    Fagerberg, L. et al. Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody-based proteomics. Mol. Cell Proteomics 13, 397–406 (2014).

  170. 170.

    Lewy, A. J., Ahmed, S., Jackson, J. M. & Sack, R. L. Melatonin shifts human circadian rhythms according to a phase-response curve. Chronobiol. Int. 9, 380–392 (1992).

  171. 171.

    Burgess, H. J., Revell, V. L., Molina, T. A. & Eastman, C. I. Human phase response curves to three days of daily melatonin: 0.5 mg versus 3.0 mg. J. Clin. Endocrinol. Metab. 95, 3325–3331 (2010).

  172. 172.

    Gerdin, M. J. et al. Melatonin desensitizes endogenous MT2 melatonin receptors in the rat suprachiasmatic nucleus: relevance for defining the periods of sensitivity of the mammalian circadian clock to melatonin. FASEB J. 18, 1646–1656 (2004).

  173. 173.

    Witt-Enderby, P. A., Masana, M. I. & Dubocovich, M. L. Physiological exposure to melatonin supersensitizes the cyclic adenosine 3ʹ,5ʹ-monophosphate-dependent signal transduction cascade in Chinese hamster ovary cells expressing the human mt1 melatonin receptor. Endocrinology 139, 3064–3071 (1998).

  174. 174.

    Owino, S., Contreras-Alcantara, S., Baba, K. & Tosini, G. Melatonin signaling controls the daily rhythm in blood glucose levels independent of peripheral clocks. PLOS ONE 11, e0148214 (2016).

  175. 175.

    Gbahou, F. et al. Design and validation of the first cell-impermeant melatonin receptor agonist. Br. J. Pharmacol. 174, 2409–2421 (2017).

  176. 176.

    Oishi, A. et al. Orphan GPR61, GPR62 and GPR135 receptors and the melatonin MT2 receptor reciprocally modulate their signaling functions. Sci. Rep. 7, 8990 (2017).

  177. 177.

    Kamal, M. et al. Convergence of melatonin and serotonin (5-HT) signaling at MT2/5-HT2C receptor heteromers. J. Biol. Chem. 290, 11537–11546 (2015).

  178. 178.

    Santos, R. et al. A comprehensive map of molecular drug targets. Nat. Rev. Drug Discov. 16, 19–34 (2017).

  179. 179.

    Hauser, A. S., Attwood, M. M., Rask-Andersen, M., Schioth, H. B. & Gloriam, D. E. Trends in GPCR drug discovery: new agents, targets and indications. Nat. Rev. Drug Discov. 16, 829–842 (2017).

  180. 180.

    Liu, J. et al. MT1 and MT2 melatonin receptors: a therapeutic perspective. Annu. Rev. Pharmacol. Toxicol. 56, 361–383 (2016).

  181. 181.

    Millan, M. J. et al. The melatonergic agonist and clinically active antidepressant, agomelatine, is a neutral antagonist at 5-HT2C receptors. Int. J. Neuropsychopharmacol. 14, 768–783 (2011).

  182. 182.

    US National Library of Medicine. ClinicalTrials.gov https://clinicaltrials.gov/ct2/show/NCT02615002 (2018).

  183. 183.

    Neurim Pharmaceuticals. Piromelatine. Neurim Pharmaceuticals http://www.neurim.com/products/piromelatine (2018).

  184. 184.

    Clarke, T. C., Black, L. I., Stussman, B. J., Barnes, P. M. & Nahin, R. L. Trends in the use of complementary health approaches among adults: United States, 2002–2012. Natl Health Stat. Rep. 79, 1–16 (2015).

  185. 185.

    Black, L. I., Clarke, T. C., Barnes, P. M., Stussman, B. J. & Nahin, R. L. Use of complementary health approaches among children aged 4–17 years in the United States: National Health Interview Survey, 2007–2012. Natl Health Stat. Rep. 78, 1–19 (2015).

  186. 186.

    Syndicat National des Compléments Alimentaires. Du marché des compléments alimentaires en France. Synadiet.org http://www.synadiet.org/sites/default/files/page/files/chiffres_cles_2016_version_pdf.pdf (2016).

  187. 187.

    Hauser, A. S. et al. Pharmacogenomics of GPCR drug targets. Cell 172, 41–54 (2018).

  188. 188.

    Trades Union Congress. Number of people working night shifts up by more than 250,000 since 2011, new TUC analysis reveals. TUC.org.uk http://www.tuc.org.uk/news/number-people-working-night-shifts-more-250000-2011-new-tuc-analysis-reveals (2016).

  189. 189.

    Striegel-Moore, R. H. et al. Exploring the typology of night eating syndrome. Int. J. Eat. Disord. 41, 411–418 (2008).

  190. 190.

    Forrestel, A. C., Miedlich, S. U., Yurcheshen, M., Wittlin, S. D. & Sellix, M. T. Chronomedicine and type 2 diabetes: shining some light on melatonin. Diabetologia 60, 808–822 (2017).

  191. 191.

    Simonis-Bik, A. M. et al. Gene variants in the novel type 2 diabetes loci CDC123/CAMK1D, THADA, ADAMTS9, BCL11A, and MTNR1B affect different aspects of pancreatic beta-cell function. Diabetes 59, 293–301 (2010).

  192. 192.

    Florez, J. C. et al. Effects of genetic variants previously associated with fasting glucose and insulin in the Diabetes Prevention Program. PLOS ONE 7, e44424 (2012).

  193. 193.

    Garaulet, M. et al. Common type 2 diabetes risk variant in MTNR1B worsens the deleterious effect of melatonin on glucose tolerance in humans. Metabolism 64, 1650–1657 (2015).

  194. 194.

    Lopez-Minguez, J., Saxena, R., Bandin, C., Scheer, F. A. & Garaulet, M. Late dinner impairs glucose tolerance in MTNR1B risk allele carriers: a randomized, cross-over study. Clin. Nutr. 37, 1133–1140 (2017).

  195. 195.

    Eze, I. C. et al. Exposure to night-time traffic noise, melatonin-regulating gene variants and change in glycemia in adults. Int. J. Environ. Res. Public Health 14, 1492 (2017).

  196. 196.

    Stoschitzky, K. et al. Influence of beta-blockers on melatonin release. Eur. J. Clin. Pharmacol. 55, 111–115 (1999).

  197. 197.

    Ying, S. W. et al. Melatonin analogues as agonists and antagonists in the circadian system and other brain areas. Eur. J. Pharmacol. 296, 33–42 (1996).

  198. 198.

    Melke, J. et al. Abnormal melatonin synthesis in autism spectrum disorders. Mol. Psychiatry 13, 90–98 (2008).

  199. 199.

    Chaste, P. et al. Genetic variations of the melatonin pathway in patients with attention-deficit and hyperactivity disorders. J. Pineal Res. 51, 394–399 (2011).

  200. 200.

    Chaste, P. et al. Identification of pathway-biased and deleterious melatonin receptor mutants in autism spectrum disorders and in the general population. PLOS ONE 5, e11495 (2010).

  201. 201.

    Reiter, R. J. et al. Mitochondria: central organelles for melatonin’s antioxidant and anti-aging actions. Molecules 23, 509 (2018).

  202. 202.

    Aschoff, J. Circadian rhythms in man. Science 148, 1427–1432 (1965).

  203. 203.

    Mayeuf-Louchart, A., Zecchin, M., Staels, B. & Duez, H. Circadian control of metabolism and pathological consequences of clock perturbations. Biochimie 143, 42–50 (2017).

  204. 204.

    Perelis, M. et al. Pancreatic beta cell enhancers regulate rhythmic transcription of genes controlling insulin secretion. Science 350, aac4250 (2015).

  205. 205.

    Ruiter, M. et al. The daily rhythm in plasma glucagon concentrations in the rat is modulated by the biological clock and by feeding behavior. Diabetes 52, 1709–1715 (2003).

  206. 206.

    Marcheva, B. et al. Disruption of the clock components CLOCK and BMAL1 leads to hypoinsulinaemia and diabetes. Nature 466, 627–631 (2010).

  207. 207.

    Saini, C. et al. A functional circadian clock is required for proper insulin secretion by human pancreatic islet cells. Diabetes Obes. Metab. 18, 355–365 (2016).

  208. 208.

    West, A. C. & Bechtold, D. A. The cost of circadian desynchrony: evidence, insights and open questions. Bioessays 37, 777–788 (2015).

  209. 209.

    Simonneaux, V. Naughty melatonin: how mothers tick off their fetus. Endocrinology 152, 1734–1738 (2011).

  210. 210.

    Muhlbauer, E., Gross, E., Labucay, K., Wolgast, S. & Peschke, E. Loss of melatonin signalling and its impact on circadian rhythms in mouse organs regulating blood glucose. Eur. J. Pharmacol. 606, 61–71 (2009).

  211. 211.

    de Farias Tda, S. et al. Pinealectomy interferes with the circadian clock genes expression in white adipose tissue. J. Pineal Res. 58, 251–261 (2015).

  212. 212.

    Sun, M. et al. Meta-analysis on shift work and risks of specific obesity types. Obes. Rev. 19, 28–40 (2018).

  213. 213.

    Anothaisintawee, T., Reutrakul, S., Van Cauter, E. & Thakkinstian, A. Sleep disturbances compared to traditional risk factors for diabetes development: systematic review and meta-analysis. Sleep Med. Rev. 30, 11–24 (2016).

  214. 214.

    Da Silva Xavier, G. The cells of the islets of Langerhans. J. Clin. Med. 7, E54 (2018).

  215. 215.

    Rorsman, P. & Ashcroft, F. M. Pancreatic beta-cell electrical activity and insulin secretion: of mice and men. Physiol. Rev. 98, 117–214 (2018).

  216. 216.

    Szewczyk-Golec, K. et al. Melatonin supplementation lowers oxidative stress and regulates adipokines in obese patients on a calorie-restricted diet. Oxid. Med. Cell. Longev. 2017, 8494107 (2017).

  217. 217.

    Chojnacki, C. et al. Effects of fluoxetine and melatonin on mood, sleep quality and body mass index in postmenopausal women. J. Physiol. Pharmacol. 66, 665–671 (2015).

  218. 218.

    Goyal, A. et al. Melatonin supplementation to treat the metabolic syndrome: a randomized controlled trial. Diabetol. Metab. Syndr. 6, 124 (2014).

  219. 219.

    Mesri Alamdari, N. et al. A double-blind, placebo-controlled trial related to the effects of melatonin on oxidative stress and inflammatory parameters of obese women. Horm. Metab. Res. 47, 504–508 (2015).

  220. 220.

    Romo-Nava, F. et al. Melatonin attenuates antipsychotic metabolic effects: an eight-week randomized, double-blind, parallel-group, placebo-controlled clinical trial. Bipolar Disord. 16, 410–421 (2014).

  221. 221.

    Cichoz-Lach, H., Celinski, K., Konturek, P. C., Konturek, S. J. & Slomka, M. The effects of L-tryptophan and melatonin on selected biochemical parameters in patients with steatohepatitis. J. Physiol. Pharmacol. 61, 577–580 (2010).

  222. 222.

    Celinski, K. et al. Effects of treatment with melatonin and tryptophan on liver enzymes, parameters of fat metabolism and plasma levels of cytokines in patients with non-alcoholic fatty liver disease — 14 months follow up. J. Physiol. Pharmacol. 65, 75–82 (2014).

  223. 223.

    Koziróg, M. et al. Melatonin treatment improves blood pressure, lipid profile, and parameters of oxidative stress in patients with metabolic syndrome. J. Pineal Res. 50, 261–266 (2011).

  224. 224.

    Borba, C. P. et al. Placebo-controlled pilot study of ramelteon for adiposity and lipids in patients with schizophrenia. J. Clin. Psychopharmacol 31, 653–658 (2011).

  225. 225.

    Kedziora-Kornatowska, K. et al. Melatonin improves oxidative stress parameters measured in the blood of elderly type 2 diabetic patients. J. Pineal Res. 46, 333–337 (2009).

  226. 226.

    Cavallo, A., Daniels, S. R., Dolan, L. M., Bean, J. A. & Khoury, J. C. Blood pressure-lowering effect of melatonin in type 1 diabetes. J. Pineal Res. 36, 262–266 (2004).

  227. 227.

    Wakatsuki, A., Okatani, Y., Ikenoue, N., Kaneda, C. & Fukaya, T. Effects of short-term melatonin administration on lipoprotein metabolism in normolipidemic postmenopausal women. Maturitas 38, 171–177 (2001).

  228. 228.

    Tamura, H. et al. Melatonin treatment in peri- and postmenopausal women elevates serum high-density lipoprotein cholesterol levels without influencing total cholesterol levels. J. Pineal Res. 45, 101–105 (2008).

  229. 229.

    Amstrup, A. K. et al. Reduced fat mass and increased lean mass in response to 1 year of melatonin treatment in postmenopausal women: a randomized placebo-controlled trial. Clin. Endocrinol. 84, 342–347 (2016).

  230. 230.

    Tsunoda, T. et al. The effects of ramelteon on glucose metabolism and sleep quality in type 2 diabetic patients with insomnia: a pilot prospective randomized controlled trial. J. Clin. Med. Res. 8, 878–887 (2016).

  231. 231.

    Walecka-Kapica, E. et al. The effect of melatonin supplementation on the quality of sleep and weight status in postmenopausal women. Prz. Menopauzalny 13, 334–338 (2014).

  232. 232.

    Gonciarz, M. et al. The pilot study of 3-month course of melatonin treatment of patients with nonalcoholic steatohepatitis: effect on plasma levels of liver enzymes, lipids and melatonin. J. Physiol. Pharmacol. 61, 705–710 (2010).

  233. 233.

    Gonciarz, M. et al. The effects of long-term melatonin treatment on plasma liver enzymes levels and plasma concentrations of lipids and melatonin in patients with nonalcoholic steatohepatitis: a pilot study. J. Physiol. Pharmacol. 63, 35–40 (2012).

  234. 234.

    Kedziora-Kornatowska, K. et al. Antioxidative effects of melatonin administration in elderly primary essential hypertension patients. J. Pineal Res. 45, 312–317 (2008).

  235. 235.

    Mostafavi, A. et al. Melatonin decreases olanzapine induced metabolic side-effects in adolescents with bipolar disorder: a randomized double-blind placebo-controlled trial. Acta Med. Iran. 52, 734–739 (2014).

  236. 236.

    Cavallo, A., Daniels, S. R., Dolan, L. M., Khoury, J. C. & Bean, J. A. Blood pressure response to melatonin in type 1 diabetes. Pediatr. Diabetes 5, 26–31 (2004).

  237. 237.

    Rindone, J. P. & Achacoso, R. Effect of melatonin on serum lipids in patients with hypercholesterolemia: a pilot study. Am. J. Ther. 4, 409–411 (1997).

  238. 238.

    Scheer, F. A., Van Montfrans, G. A., van Someren, E. J., Mairuhu, G. & Buijs, R. M. Daily nighttime melatonin reduces blood pressure in male patients with essential hypertension. Hypertension 43, 192–197 (2004).

  239. 239.

    Cagnacci, A. et al. Prolonged melatonin administration decreases nocturnal blood pressure in women. Am. J. Hypertens. 18, 1614–1618 (2005).

Download references

Acknowledgements

We thank Julie Dam and Erika Cecon (Institut Cochin, France) for their valuable expert advice during the preparation of the manuscript. The authors were supported by the Agence Nationale de la Recherche (ANR-2011-BSV1-012-01 “MLT2D” and ANR-2011-META “MELA-BETES”), the Fondation de la Recherche Médicale (Equipe FRM DEQ20130326503), Institut National de la Santé et de la Recherche Médicale (INSERM) and Centre National de la Recherche Scientifique (CNRS).

Reviewer information

Nature Reviews Endocrinology thanks J. Cipolla-Neto and other anonymous reviewers for their contribution to the peer review of this work.

Author information

Both authors contributed to all aspects of the manuscript.

Correspondence to Ralf Jockers.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Karamitri, A., Jockers, R. Melatonin in type 2 diabetes mellitus and obesity. Nat Rev Endocrinol 15, 105–125 (2019) doi:10.1038/s41574-018-0130-1

Download citation

Further reading