Investigating the relationship between melatonin levels, melatonin system, microbiota composition and bipolar disorder psychopathology across the different phases of the disease

Bipolar disorder (BD) is a complex chronic psychiatric condition characterized by recurrent episodes of depression and mania/hypomania alternating with intervals of well-being. According to the World Health Organization, BD is among the top 10 leading cause of years lived with disability (Murray and Lopez 1996). Epidemiological estimates show that 1–5% of people across all populations suffers from diverse types of BD (Angst et al. 2003). Further, BD carries a high burden in terms of treatment costs (Dilsaver 2011), effect on families and careers, loss of productivity and increased mortality (Lomholt et al. 2019). Patients with BD experience substantial residual morbidity resulting in up to 50% of follow-up time spent ill, mainly in depression (Joffe et al. 2004; Post et al. 2003), with a risk of relapse at 5 years that remains high (about 70%) even in the presence of maintenance treatment (Gitlin et al. 1995). Selecting the most effective therapy for a BD patient thus still remains a challenge for clinicians (Geddes and Miklowitz 2013) and despite progress in the study of the neurobiology and psychopharmacology of the disease, options remain limited and results are often unsatisfactory. Thus, more research is needed to better understand the neurobiology of the disease and to discover and validate novel potential targets for drug development.

A key component of BD psychopathology (i.e. illness cyclicity) appears to correspond to a neurobiological malfunction of the circadian clock system (Dallaspezia and Benedetti 2009, 2015) that is regulated by the suprachiasmatic nucleus (SCN). Patients affected by BD show alterations of intrinsic biological rhythms, including the sleep–wake cycle, as well as of hormonal rhythms and of the regulation of body temperature (Harvey 2008). A key neurochemical modulator of the circadian clock system is melatonin (MLT). This neuromodulator derives from serotonin (5-HT) and is mainly synthesized in the pineal gland during the night. The physiological effects of MLT are mostly mediated by two high-affinity G-protein coupled receptors (GPCRs), MT₁ and MT₂ (Dubocovich et al. 2010; Jockers et al. 2016). MLT is one of the endogenous entraining signals for the circadian system, regulating the activity of the SCN that expresses both MT₁ and MT₂ receptors (Dubocovich et al. 2010; Jockers et al. 2016). However, the exact mechanism/s through which MLT regulates the SNC is still to be clarified. Indeed, in vitro and in vivo studies have reported contrasting findings (Dubocovich and Markowska 2005) [for details on this topic, please see our recent papers (Gobbi and Comai 2019a, b)]. The MLT system seems to be a potential target system for the treatment of mood disorders since agomelatine, a non-selective agonist for both MT₁ and MT₂ receptors and an antagonist for 5-HT_2C receptors has entered clinical use for the treatment of major depression (de Bodinat et al. 2010).

Given the link between MLT and the circadian rhythms, and between the latter and BD (McCarthy 2019; Gonzalez et al. 2018), a series of seminal studies have investigated and demonstrated changes in the levels of MLT associated with illness status in patients with BD (Kennedy et al. 1996; Lam et al. 1990; Nurnberger et al. 2000). In particular, affected subjects have low serum levels of MLT than healthy controls (trait marker) (Kennedy et al. 1996). However, Whalley et al. (1991) did not replicate this finding. Although there is no apparent variation in the MLT levels in relation to a specific phase of the disease (Kennedy et al. 1996), the peak of MLT at night is delayed in euthymic BD patients (Nurnberger et al. 2000). In addition, Nurnberger et al. (2000) showed higher light-induced suppression of MLT in BD patients than in controls, whereas, in contrast, Lam et al. (1990) reported greater light-induced suppression of MLT in controls than in BD patients. Whalley et al. (1991) instead did not find any significant difference in the effect of light on MLT comparing healthy individuals and BD patients (Whalley et al. 1991). Therefore, the presence of a MLT supersensitivity to light as a reliable trait marker of BD needs to be clarified. Overall, the small size of the population analyzed strongly limits the validity of the currently available studies.

In addition to a possible link between dysfunctional MLT levels and BD pathophysiology, there is also a strong hint for a modulating effect of treatments for BD on MLT levels and function. Indeed, in healthy volunteers the mood stabilizer valproic acid significantly decreased the sensitivity of MLT to light (Hallam et al. 2005). Further, in a study in Long Evans rats, lithium significantly influenced MLT levels in several brain regions along the retinal-hypothalamic-pineal pathway (Seggie et al. 1987).

Another biological component, non-genetic but heritable to some extent, that could impact brain function is the gut microbiota. There is growing evidence that the bidirectional signalling along the gut-brain axis may play an important role in the pathophysiology of depression, anxiety and autism spectrum disorder (ASD) (Foster and McVey Neufeld 2013; Sharon et al. 2019; Zheng et al. 2016). Relevant to our project is the evidence that microbiota alterations could modulate the risk of BD, although findings are still preliminary (Nguyen et al. 2018; Pisanu and Squassina 2018). For instance, a diminished representation of Faecalibacterium and of an as yet unclassified member of the Ruminococcaceae family has been detected in subjects with BD with respect to controls (Evans et al. 2017). Interestingly, BD patients with higher levels of Faecalibacterium displayed better sleep quality and less severe depressive symptoms (Evans et al. 2017).

The gut is the main organ involved in peripheral synthesis of MLT, and MLT and its precursor tryptophan exert an important regulatory role on the gastro-intestinal (GI) functions (Bubenik 2002). Indeed, altered tryptophan metabolism in the gut can lead to increased GI permeability, which in turn has been associated with psychiatric disorders (Kelly et al. 2015). MLT, when administered in mice subjected to a high-fat diet (HFD), reduces body weight, hepatic steatosis, low-grade systemic inflammation, improves insulin resistance, and significantly modifies the composition of the intestinal microbiota (Xu et al. 2017). A recent study indicates that MLT improves lipid metabolism in mice with a HFD diet probably through reprogramming of the microbiota (Yin et al. 2018). These still limited and recent findings highlight the importance of the intestinal microbiota in mediating the various physiological functions of MLT.

Collectively, this evidence suggests a potential involvement of MLT and of the MLT system in the physiopathology of BD. The present study protocol is designed to explore in a sample of BD patients the possible link between the MLT system, microbiota and psychopathology across the different mood phases of the illness. In particular, we aim at exploring whether the patterns of associations identified between the levels of MLT and its precursors [tryptophan (Trp), 5-hydroxytryptophan (5-HTP), and 5-HT)] and BD mood phases are modulated by variants within the genes encoding for the elements of the MLT system, including MLT receptors and the enzymes involved in the synthesis and metabolism of MLT, and the microbiota composition.

This study protocol is composed of two approaches: (1) a cross sectional part that will test whether differences in MLT levels are associated with illness status in BD patients (case–control comparison), and (2) a longitudinal prospective analysis lasting 24 months in which recruited patients will be monitored for the manifestations of mood episodes. Clinical measures and biological material will be collected at baseline in euthymia and when a significant mood disruption will be observed (case-only analysis). In both approaches we will test whether genetic and microbiota variation can modulate the identified associations. Using this approach, we expect to find a significant difference in the plasma levels of MLT and/or some of its precursors between healthy controls and BD patients, and also differences in these biomarkers in BD patients depending on the phase of the disease (depression versus hypomania/mania versus euthymia). Importantly, we also expect to find genetic alterations, in particular in those genes coding for elements of the MLT system, modulating the longitudinal association between the MLT plasma levels and the disease morbidity in BD patients. Very likely, we will identify some typical BD symptomatic clusters that will be more strictly associated with variations in the MLT system. These data could suggest that this group of patients could benefit from pharmacological treatment acting on the MLT system. Furthermore, another important expected result will be the identification of specific microbiota clusters associated with the peripheral variations of MLT, also in relation to the BD disease phase. The combination of extensive phenotypic information with omics data can lead to the discovery of novel, potentially clinically relevant associations, especially when adequate data mining techniques are applied (Breuer et al. 2018).

The study design (cross-sectional in the comparison between controls and BD), however, will not allow us to establish a causality, i.e. if the altered MLT levels precede the development of BD. Therefore, although BD patients will be followed-up longitudinally, this study can be only considered as a proof-of-concept, and thus as a hypothesis generator and identifier of the size of the association between the variables under exam. In spite of these limitations, this study should be able to analyze, in depth and at different levels, the role of the MLT system (MLT and its precursors’ levels and the genetic variability) and its correlation with the microbiota composition in BD pathophysiology. Another limitation may arise by the fact that we will not study the entire 24-h secretion pattern of MLT and the possible variation in its precursors (e.g. by measuring their levels at different time points during the day and the night or measuring the 24-h excretion of 6-sulphatoxymelatonin in the urine) in relationships with the different biological and psychometric markers.

Collectively, these data will help us to better clarify the neurobiology of BD and will be helpful for future projects aimed at the development of new drugs that, by targeting the MLT system, could help improving some of the symptoms of BD. It is important to remember that, to date, there are several pharmacological treatments for BD, but the rate of recurrence of the disease and the toxicity of these drugs are significant. Therefore, there is a great scientific and clinical interest in the development of new drugs, a process for which it is first necessary to better understand the neurobiology of BD.