In BD type 1, patients experience at least one manic episode and various depressive episodes, whereas in BD type 2 they undergo at least one depressive episode and at least one hypomanic episode and have no history of manic episodes. A patient with type I or type II BD with “mixed features” consist of the co-occurrence either of hypomanic/manic episodes with 3 or more depressive symptoms or depressive episodes with 3 or more hypomanic/manic symptoms.

Epidemiologically, systematic review and meta-analysis found that BD type 1 has a pooled lifetime prevalence of 1.06% and 1.57% for BD type 2 [10], but numbers keep rising. On the other hand, in the last years, large sample studies have been analyzed reporting higher frequency in female patients in both types of BD [11, 12]. These studies argue that there must be sex differences as women have reported to present rapid cycling more often than men, besides the onset in female patients occurs with depressive phases and present more mixed manic presentations Moreover, BD can be accompanied by psychiatric comorbidities such as anxiety or substance abuse and impulse control disorders [13]. Not only BD may have devastating impact on patients but also an important economic burden for healthcare systems, with a range of direct costs per patient and per year oscillating from $881 to $27,617 (plus $1568–$116,062 indirect costs) [14]. On the other hand, this disease also involves high social stigma; one in four patients with BD reports high internalized stigma [15]. In relation to this fact, suicide affects 23–26% of BD patients, which represents about 3.4–14% of all suicide deaths [16], and sex-specific prevalence in this sense has also been observed, being relative suicide attempts higher in bipolar women, although less violent than men [11].

Nowadays, due to the complexity of this psychiatric disorder, there is neither molecular biomarker nor biological sign used for the diagnosis of BD in the clinical routine, so this entity remains a descriptive syndrome whose diagnosis is eminently clinical, based on the already mentioned clinical manuals such as the DSM-V and the ICD [17]. The pathophysiological bases of BD are not fully understood and deepen on the biological mechanisms underlying this intricate malady is an imperative need, as this could open potential translational applications.

What is the pathophysiological basis of bipolar disorders?

To date, the etiopathogenesis of this disorder remains uncompleted and unclear, being plausible the hypothesis about the conjunction of both genetic, environmental and psychosocial factors. The heritability of BD seems to obey a complex non‐Mendelian inheritance [18] and some risk variants have been identified in previous works [19]. Thus, more than genes of major effect, evidence suggest that there are multiple susceptibility loci, each one with small effect. Furthermore, these loci seem to overlap with other psychiatric disorders, specially schizophrenia (SZ) [20]. Regarding environmental factors, compelling evidence support that they may influence the development of many psychiatric disorders, including BD. There is circumstantial evidence that certain environmental factors in pregnancy or in adulthood like viral infections could be associated with the clinical course of BD, although more evidence in this sense is warranted [21]. In the interface between genetics and environmental factors, epigenetics can be considered as the major biological mechanism explaining the etiopathogenesis of many psychiatric disorders [22, 23]. There are different epigenetic mechanisms involved in the regulation of DNA expression, including DNA methylation, histone methylation/acetylation and other modifications, along with non-coding RNAs such as micro RNAs (miRNAs) or long non-coding RNAs (lncRNAs) [24]. In the event of BD, obtained results are more limited than in other psychiatric disorders, although some authors argue that because of the nature of BD and its therapy, the field of epigenetics can be critical for understanding the clinical course of this complex disorder, aiding to define the role of environmental factors, and the possible identification of promising state and traits biomarkers [25].

From a molecular level, alterations in calcium signaling appear to be clearly associated with the development of BD [20, 26]. The intracellular calcium is essential for the modulation of several intracellular signaling cascades and neurotransmitter release [27]. Indeed, neurotransmitter dysregulation strongly underlies the pathophysiology of BD. Several neurotransmitters must be mentioned here, including monoamines (serotonin, dopamine and norepinephrine), glutamate, acetylcholine and gamma-aminobutyric acid (GABA). Importantly, the levels of these neurotransmitters can be different according to the phase of BD (Fig. 2) and can partially explain the cyclic fluctuations observed in these patients [28,29,30,31,32,33,34,35]. The aberrant neurotransmission is associated with several structural and functional changes in the brain and other encephalic structures, as it has been observed that patients with BD display abnormalities in neural circuitries supporting emotion processing, emotion regulation and reward processing [36]. These changes can also be different according to the history of these patients. For instance, image studies have found significant differences in brain activation of psychotic versus non-psychotic BD patients, showing that the former appears to be related to altered activation in left-sided regions whereas the latter exhibited right-sided functional alterations [37]. The brain of patients with BD are also characterized by impaired neurogenesis and neuroplasticity [38, 39]. In this sense, peripheral variations in some critical molecular markers involved in these processes such as brain-derived neurotrophic factor (BDNF) can be applied to understand the significant changes occurred in the brain, showing promising translational applications [40]. Alterations in different neuropeptides such as neuropeptide Y (NPY) and somatostatin have also been observed in patients with BD [41].

Main factors described are differences in genetics (especially risk variants), epigenetics (DNA and histone modifications, non-coding RNAs), metabolic alterations; aberrant calcium signaling; alterations in circadian rhythms; oxidative stress, characteristic low-grade chronic inflammation, HPA axis hyperactivation, related to psychological stress and different psychosocial factors, accompanied with multiple changes in different neural circuitries, aberrant neurotransmission and altered neuropeptides. Peripheral blood concentrations of neurotransmitters are altered in BD, where in each phase there seems to be a different pattern of activity for every neurotransmissor. These differences are relevant with respect to healthy controls. Many of these events, together with additional mechanisms like altered biological clocks, mitochondrial dysfunction or telomere shortening can be considered as signs of premature aging. Beyond, all these factors influences and are directly influence by the dysregulation of the MGB axis which acts as a central element of psychiatric disorders. HPA hypothalamic–pituitary–adrenal, GABA gamma-aminobutyric acid.

In addition, the role of oxidative stress markers in BD has been explored, as it seems that numerous lines of investigation of BD pathophysiology converge on oxidative stress and aberrations in oxidative energy generation. Although findings are not always steady, an increase in lipid peroxidation and in nitric oxide levels have been reported in BD patients, as well as increased damage in deoxyribonucleotide acid (DNA) and ribonucleic acid (RNA), compared with healthy controls [42]. Similarly, patients with BD exhibit telomere shortening which in turn can be associated with oxidative damage and other pathogenic mechanisms involved in BD [31]. Oxidative stress is also closely linked to mitochondrial dysfunction and despite findings are not consistent, it seems that deleted mitochondrial DNA is a manifestation in postmortem brains of patients with BD, which can be equally related to calcium dysregulation [43]. Even alterations in circadian rhythms seem to be implicated in BD [44], and it is thought that this fact potentiate episodes of mania and depression [45]. The dysregulation of sleep/wake cycle mediated by a delayed melatonin hormone onset reveals in most cases an eveningness chronotype [46] in these patients.

Similarly, hyperactivation of the hypothalamic–pituitary–adrenal (HPA) axis and aberrant glucocorticoid signaling is critically associated with this disorder. Although there is disagreement in affirming it is an etiological factor, what is undeniable is that this is a contributor to the BD clinical presentation and increases the risk of cognitive deterioration [47]. The hyperactivation of the HPA axis is closely related to the response of an individual to different psychosocial stressors. Among some of the most relevant psychosocial factors potentially related to BD, it must be highlighted cognitive dysfunction, altered domains of social cognition (theory of mind, emotion comprehension, empathy), autobiographical memory, temperament and personality factors (i.e. ruminative tendencies and neuroticism), which may drive to several difficulties in familial and social relationships [48]. Furthermore, there is a significant low-grade chronic inflammation reported in patients with BD, with an aberrant immune activation observed in the gut and systemically, which influences HPA axis, gut microbiota, metabolic functions, and the brain, driving to a phenomenon of neuroinflammation [49]. Likewise, dysregulation in the brain and systemic metabolism together with an altered hormonal profile represents another major feature of BD [50, 51].

As shown, the many factors involved in the etiopathogenesis of BD cannot be understood separately, but collectively exert a synergic effect, interacting bidirectionally. Beyond, many of the observed changes occurred in patients with BD can be considered as signs of premature aging (telomere length, oxidative stress, inflammation, disruptions on biological clocks, epigenetic aging and mitochondrial dysfunction), as prior works hypothesize that accelerated aging process is potentially implicated in the development and course of BD [52]. Likewise, another key branch to understand BD pathogenesis is the disruption of the so-called microbiota–gut–brain (MGB) axis [53,54,55]. Hence, the MGB axis can be considered a part of a great picture that influences and in turn is influenced by the different mechanisms involved in BD pathobiology. Because of that, there is a growing claim to study the gut microbiota and MGB axis as a potential biomarker for BD patients, offering promising clinical and translational opportunities [56]. Overall, a global picture of the plenty biological mechanisms involved in the etiopathogenesis of BD is summarized in Fig. 2.

How is operating the microbiota–gut–brain axis in bipolar disorders?

Gut microbiota typifies a complex ecosystem consisting of trillions of microbes that inhabit the human intestine and which maintain a symbiotic relationship with their host [57]. This ecosystem represents about 2% of our body weight (approximately 1.5 kg) [58]. On the other hand, the available studies are also focused on gut microbiome, which refers not only to the different microbial populations that inhabit in the gut, but also their genome and end-products [59]. Bacteria are the most important component in gut microbiota, but also viruses—especially phages—Fungi, Amoebozoa or Archaea can be found. Among bacteria, Firmicutes, Bacteroides, Proteobacteria and Actinobacteria phyla represent over 90% of the gut microbiota community [60]. The shaping of gut microbiota starts at birth. The composition of gut microbiota differs according to delivery mode, so vaginal birth is preferable to cesarean section, which is linked with gut microbiota dysbiosis and even with subsequent repercussions in immunological and metabolic status [61, 62]. Successively, the gut microbiota is highly sensitive to environmental signals, receiving, integrating and responding to the information not only from the different organs of the body, but also from external influences like diet, physical activity, psychological and physical stress, sleep restrictions, socioeconomic status, drugs, antibiotics, exposure to pets, noise, and temperature [63]. In this sense, gut microbiota signifies a field of growing interest in the understanding of several processes in the organism regulating a plethora of metabolic routes, having a very close interplay with the immune system [60]. Because of the many functions that the gut microbiota fulfills, compelling evidence suggest that gut microbiota dysregulation could be directly associated with low-grade inflammation and a wide range of pathological conditions including obesity and metabolic disorders like type 2 diabetes [64, 65], inflammatory bowel disease (IBD) [66], gastrointestinal cancer [67] and psychiatric disorders, including BD [68].

There are several ways by which MGB axis works. A complex interplay between the gut microbiota, the parasympathetic nervous system—with a prominent role of vagus nerve-immune system and the different cells located in the gut is continuously occurring [69]. These interactions can occur by direct contact or indirectly through the secretion of a plenty of specific products and metabolites which influence the different parts of the MGB axis and the whole organism [70, 71]. The mucus layer of the gut is the scenario where most of the host-microbiota interactions take place. In these interactions, there is a basal and physiological inflammatory environment induced by gut microbiota that allows the regulation of bacterial populations, preventing its spread [71]. In this context, enterocytes communicate through innate immune receptors, chemokines and cytokines. For instance, Toll-like receptors (TLR) enable host innate immune system to identify pathogen-associated molecular patterns (PAMP) of microbes, such as lipopolysaccharide (LPS), lipoproteins or flagellin [72]. The brain receives and integrates signals from the gut directly from the afferent fibers of the vagus nerve [73] or due to the aforementioned products secreted by the different cells to the systemic circulation. In turn, the brain exerts a direct influence on the MGB axis through the efferent fibers of the vague nerve, or indirectly (i.e., by the activation of the HPA axis), denoting the bidirectional interplay occurred in this axis [74].

In this attractive and promising background, the aim of the present work is to study the relationships between BD and gut microbiota in the context of pathophysiology, and how the MGB axis interacts with the multiple systems and mechanisms behind this psychiatric condition. Equally, we will also collect the most relevant insights on the impact of BD pharmacotherapy on the gut microbiota, reviewing promising translational approaches targeting the MGB axis as well.

As shown, BD patients are characterized by presenting a reduced microbial diversity and changes in relative abundance of several taxa, which may be involved in the pathogenesis of this complex disorder. Besides, specific differences according to the type and clinical manifestation, although further efforts are needed in hypomanic/manic phases, mixed features and also in the comparison between type 1 and 2 BD. Finally, compelling evidence is also supporting the notion that the gut microbiota can present some similarities and disparities across psychiatric disorders, opening the use of gut microbiota as potential biomarkers.

On the other hand, there are also some studies evaluating the gut microbiota composition in relation to the type of BD. Interestingly, there are works focused on comparing gut microbiota composition in patients with MDD, type I and type II BD. Notwithstanding there are some similarities in the MGB axis disruption in these three groups in comparison to healthy subjects, it seems that these aberrations are greater in type I BD patients and melancholia [56]. Indeed, individuals with type I BD showed higher IgM responses to Morganella morganiithan patients with MDD and type II BD. Similarly, McIntyre et al. [82] reported that patients with type II BD exhibit a greater abundance of Collinsella in comparison to those with type I, although they noticed about the small sample size and insufficient control for some potential moderating factors like medication. More studies could be aimed to find if there is a differential gut microbiota profile in patients with type 1 versus type 2 BD (Fig. 3).

Also, some researchers have found in gut microbiota a promising point to compare and distinguish between BD and other psychiatric disorders. For instance, it seems that Prevotella 2 and Ruminococcaceae UCG-002 are more prevalent in patients with MDD than BD patients [83]. Likewise, compared with healthy individuals, MDD is associated with alterations in Bacteroidaceae family, whereas disturbances in Lachnospiraceae, Prevotellaceae, and Ruminococcaceae families are related to BD [84]. Furthermore, the abundance of Fusobacteriaceae, Escherichia blattae DSM 4481 and Klebsiella oxytoca were significantly augmented, whereas the Bifidobacterium longum subsp. infantis ATCC 15697 = JCM 1222was significantly reduced in BD group compared with MDD group [85]. Besides, a very recent systematic review compared the gut microbiota in BD, MDD and SZ [86]. Interestingly, they observe that an increase in Eggerthella and Lactobacillus, together with a decrease in Coprococcus was common for the three conditions in comparison to controls. Likewise, MDD and SZ shared higher levels of Escherichia/Shigella and Veillonella whereas increased Megasphaera and lower Roseburia are common for BD and SZ. MDD and BD had more commonalities, including higher Enterococcus, Flavonifractor, and Streptococcus, and lower Faecalibacterium and Ruminococcus. They reported that each mental disorder could be potentially characterized by a different microbial profile: MDD was often characterized by higher Alistipes and Parabacteroides and lower Prevotella; BD by higher Bifidobacterium and Oscillibacter; and SZ by higher Prevotella and lower Bacteroides, Haemophilus, and Streptococcus [86]. Other studies have also found that BD and SZ are characterized by higher serum antibody levels to fungal pathogens Saccharomyces cerevisiae and Candida albicans and how these changes could be related to cognitive performance or the onset of psychotic symptoms [87].

Collectively, as shown in Fig. 3, these studies support the notion that there is an altered gut microbiota profile in patients with BD when compared to healthy subjects, and that changes observed in gut microbiota can correlate with the clinical manifestations, including both types and phases. Besides, by the analysis of gut microbiota in different psychiatric disorders the discovery of potential biomarkers that aid in the clinical diagnosis, prognosis or therapy prediction can also be opened. Due to the promising implications related to this field, we encourage further and more precise works to evaluate and study the gut microbiota in BD and other mental disorders.

Leaky gut promotes bacterial translocation and consequent endotoxemia, what is a major contributor of inflammation, locally (intestinal) and systemically. Oxidative stress is also represented, and it seems to be even increased in certain phases; particularly, in mania, an increase in dopamine causes higher ROS production. Besides HPA axis hyperactivation is related to the aberrant control release in addition to psychological stress. Chronic inflammation causes impairment at BBB integrity, and then glial activation (microglia and astrocytes) is triggered. These mechanisms cause damage to oligodendrocytes, which causes imbalance at myelinization and therefore changes at WM. This cascade of events causes the so-called neuroinflammation, which promotes clinical manifestations of BD like cognitive impairment and negative emotion reactivity. ROS reactive oxygen species, BDNF brain-derived neurotrophic factor, BBB blood–brain barrier, WM white matter, HPA hypothalamic–pituitary–adrenal, GM gut microbiota, LPS lipopolysaccharide.

Particularly, numerous BD patients manifest HPA axis hyperactivity in whole circadian rhythm, especially during mania [47]. Childhood trauma has been demonstrated to have a crucial role in HPA axis alterations in BD patients [159]. These patients showed to have increased levels of cortisol as well as lower glucocorticoid receptor’s function and alterations in glucocorticoid signaling. The hyperactivity of HPA axis is not the etiological factor that triggers BD, but undoubtable, it contributes to modify clinical presentations of BD and influence the genetic and environmental basic interaction in BD etiopathogenesis [47]. However, in patients with BD, neuropsychological performance is closely related to cortisol [160], and higher levels of this hormone have been related to poorer performance in some cognitive tasks [161, 162], especially working memory [163]. Moreover, abnormalities in this axis and in cortisol levels can be related to higher cardiovascular risk [164, 165], which may explain the increase in the mortality rate and in the cardiovascular risk factors prevalence observed in affective disorders, including BD [166].

Gut microbiota has bidirectional communication with HPA axis: certain microbial metabolites can attenuate it whereas translocated microbial antigens and exerted cytokines or prostaglandins productions activate it [95]. Moreover, dysbiosis status is associated with a constant increase in corticosterone synthesis in ileum, and, consequently, to glucocorticoid and insulin resistance, hyperglycemia and dyslipidemia [167]. This hypercortisolism is also associated with reduction in the immune response [97]. Then, HPA axis has an important impact on gut microbiota composition and neuroenteroendocrine functions, increasing gastrointestinal permeability, and contributing to the chronic low-grade inflammation [168]. Gut microbiota is not only a key player in regulating intestinal permeability, inflammation and neuroinflammation, but also in modulating behavior acting as an endocrine organ and modulating neural pathways through MGB axis as commented below.

Neuromodulatory actions of gut microbiota in bipolar disorder

Gut microbiota can have important neuromodulatory activities either by direct actions (synthesizing or degrading neurotransmitters) or indirectly through the production of different microbial metabolites, which exert a plethora of systemic actions [169]. Besides, enteric and CNS can also influence neuromodulation actions of gut microbiota. In this section we will summarize the neuromodulatory implications of gut microbiota in patients with BD, focusing on its direct actions on neurotransmission and the main microbial metabolites (SCFAs, tryptophan metabolites).

Direct actions of gut microbiota on neurotransmission in bipolar disorder

As previously described, neurotransmitter dysregulation is a major pathophysiological signature of BD. Especially, GABA, glutamate, serotonin, dopamine, norepinephrine and acetylcholine appears to be importantly dysregulated in these patients [31, 170]. Gut microbiota is a central modulator of several neurotransmitters, being able to synthesize them and their precursors as well as promoting their degradation or metabolization to other products [171]. There are some species of bacteria grown in culture that can produce serotonin, although it seems that the most prominent role of gut microbiota in serotonin synthesis is achieved through the stimulation of enteroendocrine cells (EECs), especially by spore-forming bacteria [172]. The interplay between gut microbiota and EECs mostly occurs through different microbial metabolites, with a central role of SCFAs and tryptophan, as it will be later discussed.

Glutamate and GABA can be obtained from dietary sources or produced by the gut microbiota. Several bacteria are able of producing glutamate including Coryneform and lactic acid bacteria (LAB). Interestingly, glutamate also exert important modulatory effects on EECs, sending sensory signals to the vagus nerve [173] GABA is biosynthesized from glutamate due to the action of the enzyme glutamate decarboxylase, which is found in eukaryotic cells and in a broad spectrum of bacterial species [174]. However, LAB and more specifically certain Lactobacilli strains are the most important GABA producers in the gut [175]. In this sense, one study explored the possible role of Lactobacillus and Bifidobacterium in patients with BD. Interestingly, they show that notwithstanding the counts of Bifidobacterium or Lactobacillus were not significantly varied in patients with BD, it seems to exist a negative correlation between Lactobacillus and Bifidobacterium counts with sleep and cortisol levels respectively, being both important mechanisms in BD pathophysiology [77].

A broader number of bacteria have been reported to be able to produce other monoamines like dopamine and norepinephrine whereas only Lactobacillus plantarumseems to take part in the synthesis of acetylcholine [176, 177]. Despite the role of gut microbiota in the modulation of dopamine and norepinephrine remains to be fully understood, germ-free (GF) animal models have shown that there is a noteworthy alteration of these neurotransmitters in the gut as well as in the brain, which could have important consequences in the behavior [178].

Last but not least, nearly 117 types of bacteria have been identified as major histamine producers [179]. Apart from its well-known immunomodulatory role, histamine can act as a neurotransmitter in the brain, acting as a critical modulator of other neurotransmitters, and influencing arousal, motivation, and energy balance [180]. Neurons located in the posterior hypothalamus projects to all the major regions of the CNS, also participating in the regulation of sleep-wakefulness [181]. Likewise, the central histamine system in the brain is mediated via G-protein-coupled H1-H4 receptors and also appears to be involved in additional functions like arousal, control of pituitary hormone secretion, suppression of eating and cognitive functions [182]. Currently, there are some hypotheses supporting that patients with BD may have an altered histaminergic system, with upregulated histamine levels in manic phases and downregulated in depression, being their fluctuations closely related to other pathophysiological mechanisms [183]. An upregulated histamine production in the brain seems to drive to sleep-wake alterations in vivo, being associated with behavioral and metabolic disorders similar to those caused by voluntary sleep restriction in humans [184]. Conversely, animal models have also established that low levels of histamine-induced depression-like behavior, decreased locomotor activity in the home cage, and impaired aversive memory, leading to a decrease in wakefulness as well [185]. Hence, it is likely that the different alterations reported in the gut microbiota of patients with BD can be partially involved in the differential production of histamine and in turn this may be related to the described effects. Similarly, histamine production from gut microbiota appears to provide anti-inflammatory effects by suppressing TNF-α [186]. Thus, compelling evidence suggest the relevance of gut microbiota as pivotal player of brain function, influencing in the levels of several neurotransmitters and neuromodulators [187].

Short-chain fatty acids

SCFAs are end-products of saccharolytic fermentation of indigestible carbohydrates from our diet [188]. Also, proteins and peptides can be metabolized in the cecum and colon [189], as protein fermentation lead to branched SCFAs production [188]. The concentration of SCFAs is higher in cecum and proximal colon, and their levels decrease toward the distal colon [189]. The most important SCFAs are butyrate, propionate, and acetate, accounting approximately for 80% of the total, although the production of others like formate and valerate should also be highlighted [188,189,190]. Acetate and propionate are mainly produced by members of Bacteroidetes whereas butyrate is mainly synthesized by Firmicutes. Furthermore, butyrate can also be synthesized by the gut microbiota from acetate and lactate, although the last is not considered a SCFAs [191]. Notwithstanding mostly SCFAs uptake, signaling and functions occurring in the gut, associative studies show that SCFAs have an important role in human metabolism, the immune system and the entire organism, being able to cross the BBB [192, 193]. Critically, SCFAs are master epigenetic regulators, acting as inhibitors of the enzyme histone deacetylase (HDAC), which is involved in the histone modification [194, 195]. Thus, SCFAs are pleiotropic components produced by the gut microbiota involved in multiple processes of health and disease.

In this great context, butyrate is the most widely studied SCFA. Patients with BD show reductions in butyrate-producing bacteria which, being a central feature of MGB axis disruption [68]. In this sense, lower levels of Coprococcus, Faecalibacterium and Roseburia have been reported, while there seems to be an increase in bacteria populations that consume lactate, such as Megasphaera [86]. A recent systematic review has shown that despite the need for further studies, changes in butyrate-producer bacteria like Faecalibacterium may characterize BD in both a trait and state-dependent fashion [196]. In the gut, colonocytes consume butyrate locally, and this SCFA is crucial for maintaining the intestinal barrier [197]. SCFAs also stimulate enteroendocrine cells (EECs), promoting serotonin production in the colon [198]. This augmented production of serotonin leads to enhanced levels of serotonin in systemic circulation and in the brain, also stimulating vagal pathways and similar events occur with GABA, glucagon-like peptide 1 (GLP1) and peptide YY (PYY) [190].

Besides, immune cells have a high expression of SCFAs receptors, and it seems that SCFAs can regulate the function of colonic Treg cells [189, 199]. Despite butyrate and propionate seem to have the greatest immunomodulatory properties, all SCFAs can regulate the immune response [200]. Mostly, SCFAs exert anti-inflammatory properties. They induce the production of IL-10, closely related to intestinal homeostasis [201], and the production of IL-22 by innate lymphoid cells (ILC) requires the presence of gut microbiota IL-22 is not only crucial in intestinal homeostasis and mucosal barrier function, but also in host defense against infections [202]. Interestingly, most of the pharmacological treatment used in BD showed to increase IL-22, so the efficacy of these drugs may also be in relation with changes in this immune altered response [203]. Thus, reduced levels of butyrate and other SCFAs in patients with BD can have important consequences in the intestinal inflammation and permeability.

Moreover, SCFAs also have direct effects on the brain. For instance, SCFAs are associated with improved neuronal and cognitive function, BBB integrity and decreased neuroinflammation in the brain, representing a critical mechanism of interplay between gut microbiota and glial cells [190, 204]. Hence, despite more investigation is needed, it is conceivable that SCFAs can play a key role in the neural alterations observed in psychiatric disorders. The role of SCFAs in BD symptomatology has not been studied yet. However, a recent work analyzed SCFAs in fecal samples from 125 patients with psychiatric disorders (including 23 with BD). Interestingly, they found that self-reported depressive symptoms were positively associated with fecal acetate concentrations and negatively associated with butyrate and propionate levels [205]. Furthermore, there are in vivo studies showing that SCFAs can alleviate repeated psychosocial stress alterations, exerting antidepressant and anxiolytic effects [206]. Moreover, the therapeutic use of butyrate has been demonstrated to revert manic-like behavior in rats and regulates the antioxidant enzyme activities, protecting the brain against oxidative damage [207] therefore supporting the important role that SCFAs may play in the brain alterations in patients with BD.

Overall, there are plenty of potential effects of SCFAs in patients with BD, although more studies are required to unleash the microbial populations involved in this dysregulation. Furthermore, studying local and systemic levels of SCFAs would be of great aid to understand the pathophysiological basis of MGB axis disruption in BD, especially focusing on its different phases and types.

Tryptophan-kynurenine metabolism

Tryptophan is an essential amino acid with multiple metabolic fates and crucial in cell danger response, being again gut microbiota a modulator of its metabolic pathways [208]. Indeed, differences in tryptophan metabolizing bacterial pathways can be found in patients with neurological diseases [209] and psychiatric disorders including BD [210].

Gut microbiota can play a pivotal role in the metabolism of tryptophan. Five bacterial phyla including Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, and Fusobacteria can participate in tryptophan metabolism, being those belonging to the Clostridium, Burkholderia, Streptomyces, Pseudomonas or Bacillus genera the most important modulators [209]. First, via hydroxylated pathway tryptophan is the precursor for serotonin or melatonin [176]. Thus, a large amount of serotonin is produced by enterochromaffin cells of the gastrointestinal tract [58, 208] and it is also present in enteric nerves [208]. Although tryptophan can cross the BBB, serotonin produced in the gut cannot [211]. However, serotonin can affect both vagus nerve and BBB permeability [69], and it modulates intestinal inflammation [212] and numerous physiological processes, including gastrointestinal secretion and peristalsis, vasoconstriction, behavior, and other neurological functions [177]. Besides, it can indirectly affect central serotoninergic pathways by modulating tryptophan and tryptamine availability. Thus, gut microbiota can play a central role in the serotoninergic dysfunction observed in patients with psychiatric disorders, although further studies are required.

The role of tryptamine in the brain is not well understood, although it is hypothesized that they may cross the BBB and exert important neurotransmitter or neuromodulatory actions [213]. Besides, increased tryptamine production can be associated with decreased circulating tryptophan and serotonin synthesis in the brain and could represent one mechanism by which tryptamine influence behavior [214]. Furthermore, tryptophan can be metabolized by gut microbiota to form indole and their derivates including indole-3-aldehyde, indole-3-acetic-acid and indole-3-propionic acid [176]. The role of tryptamine, indole and their derivate has been suggested in different psychiatric disorders, although little is known about their possible implication in BD [215].

On the other hand, tryptophan can be metabolized via the indoleamine 2,3-dioxygenase (IDO) [58, 176] or through the enzyme tryptophan-2,3-dioxygenase (TDO) to form N-formylkynurenine (Kyn) [216]. This is known as the Kyn pathway, and accounts for ~95% of dietary tryptophan degradation [217]. The indoles and their derivates can also be transformed into Kyn [58, 218]. Many components of the Kyn pathway are neuroactive, and influence neuroplasticity and/or exert neurotoxic effects; and this pathway is modulated by, and in turn modulates many other systems that are commonly disrupted in psychiatric disorders, including immune, endocrine, metabolic, and hormonal systems [219]. In a simple manner, Kyn can be metabolized either to kynurenic acid (KYNA), with neuroprotective properties, or to the neurotoxic components quinolinic acid (QA) and 3-hydroxykynurenine (3-HK) [208]. KYNA acts as an NMDA receptor antagonist or through the inhibition of the α7 nicotinic acetylcholine (α7nACh) receptors whereas QA acts as an agonist toward the NMDA receptors, thus promoting excitotoxic neuronal damage [220]. Regarding 3-HK, this molecule seems to exert a dual role in the CNS. On the one hand, it appears to induce oxidative damage and cell death, and high levels of 3-HK are associated with several psychiatric disorders. In contrast, some experimental studies have provided evidence of antioxidant and scavenging properties inherent to this component [221]. Howsoever, the synthesis of 3-HK from Kyn by the flavoprotein kynurenine-3-monoxygenase (KMO) is another critical point in the Kyn pathway, determining the balance between the neurotoxic component 3-HK and the neuroprotective KYNA [216]. Besides, it seems that Kyn and its downstream products exert pivotal immunomodulatory actions in the brain [222]. Thus, alterations in the Kyn pathway, with an augmentation of neurotoxic components like QA at the expense of neuroprotective derivates like KYNA, seem to have a central role in neuroinflammation in many neuropsychiatric disorders [223]. Likewise, systemic inflammation and glucocorticoids can also cross the BBB and influence the formation of QA in the brain by the microglia and infiltrated macrophages, enhancing the neuroinflammatory cascade [224]. Moreover, Kyn pathway may critically influence glutamate neurotransmission [225]. Thus, the gut microbiota can be a major modulator of Kyn pathway in different mental disorders, impairing tryptophan metabolism and influencing the brain and systemic inflammation [226]. Indeed, Kyn pathway seems to be abnormally activated in BD patients [227, 228]. Significant changes in different molecules of the Kyn pathway in BD patients have been described. For instance, meta-analysis on the peripheral blood levels of these components demonstrates that individuals with BD present lower peripheral blood levels of tryptophan, Kyn, KYNA, xanthurenic acid (a component derived from 3-HK), KYNA/Kyn and KYNA/QA ratio [229]. Besides, it seems that individuals with manic episode showed the greatest reductions in tryptophan levels whereas KYNA levels were more notably reduced among individuals in the depressive phase [229]. Other meta-analyses, however, despite showing that patients with BD present lower peripheral levels of tryptophan, Kyn or KYNA, did not find any significant differences between manic and depressed phases [230]. Intriguingly, it seems that this shift in the tryptophan metabolism from serotonin to the kyn pathway is associated with BD, MDD and SZ, but only in mood disorders (BD and MDD) there was a preferential metabolism of Kyn to the potentially neurotoxic QA [231], demonstrating the biological differences in this pathway between various psychiatric disorders. Interestingly, Kyn/tryptophan relation, which appears to represent IDO activity, is also elevated in BD patients [232]. Importantly, the increased IDO activity is a crucial mechanism related to the inflammation-induced depressive-like behavior by endotoxemia [233], thus showing the relevance that gut dysbiosis and bacterial translocation may have on BD patients. Van den Ameele et al. [234] observed that there was a strong relation between TNF-α and Kyn, Kyn/tryptophan, 3-HK and QA in the manic subgroup whereas in the depressed subgroup the KYNA/3-HK decreased and there was a strong association between interferon-y and Kyn pathway activation. Besides, both depressive and manic subgroups were characterized by presenting low levels of KYNA in comparison to healthy controls, which seems to support the relevance of the Kyn pathway in the context of neuroinflammation in BD and how it may be differentially modulated. Beyond, it seems that there is an inverse relationship between Oscillibacter (a genera increased in patients with BD) and the levels of tryptophan and KYNA, thereby supporting the relevance of gut microbiota in the altered Kyn pathway and neuroinflammation [235]. These findings could have important therapeutical implication for BD patients; as some antidepressant like ketamine can influence the Kyn pathway [227].

Taken together, this evidence suggests that gut microbiota plays a key role in BD. Gut microbiota is responsible for the disruption of intestinal permeability, which results in an upregulation of inflammatory cytokines. The gastrointestinal tract and brain maintain a crosstalk connection in which gut microbiota and inflammatory response are crucial, and this immune disruption is inextricably related to a neuroinflammatory response mediated by abnormalities in microglia and mitochondrial functions, as well as an increase in oxidative damage. CNS injury is a cause of inflammation, but this inflammation activates also autonomic nerves, including vagus nerve, feeding back this detrimental inflammation system. Aberrant neuroendocrine and immune responses seem to be critical in the precipitation and the feedback of neuroinflammatory-related disorders, such as BD. Neuromodulatory actions herein expressed in the section “Neuromodulatory actions of gut microbiota in bipolar disorder” are summarized in Fig. 5.

There are direct (synthesis or degradation of neurotransmitters and their precursors) and indirect (synthesis of microbial metabolites with systemic pleiotropic effects) actions. In the indirect actions the main metabolites, SCFAs and tryptophan (trp) derived metabolites, are represented. Red arrows represent the result of pathological condition of BD, meaning the increased or decreased level in certain metabolic routes. Decreased levels of butyrate and propionate besides increased levels of acetate have been related to higher depressive symptoms. Tryptamine levels are risen whereas tryptophan levels are decreased in BD. Kynurenine pathway is also altered in BD and it seems to be related to neuroinflammation, bacterial translocation, HPA dysfunction, neuromodulation. As shown, a peripheral decrease in many metabolites of the Kyn pathway is shown, with specific differences in manic versus depressive stages and inversely correlated with certain gut microbiota populations (Oscillibacter genera). Glutamate concentration, abnormal SCFAs production and decreased trp, all confluence in less enteric serotonin (5HT) production due to decreased stimulation of enteroendocrine cells (EECs). All these mechanisms disruption contribute to bidirectional intestinal inflammation-neuroinflammation. Central nervous system (CNS) also sends signals activating vagus nerve. Studying local and systemic levels of SCFAs in patients with BD can be of aid to better understand microbiota–gut–brain (MGB) axis disruption in this malady. More studies focusing on its different phases and types are needed. This scenario is just the beginning of new therapeutical approaches. GM gut microbiota, DA dopamine, AC acetylcholine, NE norepinephrine, his histamine, GABA gamma-aminobutyric acid, EECs enteroendocrine cells, 5HT serotonin, SCFAs short-chain fatty acids, trp tryptophan, kyn kynurenine, KYNA Kynurenic acid, XA Xanthurenic acid, QA Quinolic acid, Tregs T regulator cells, ILC innate lymphoid cell, BBB blood–brain barrier, CNS central nervous system; (+): activates.